Uranium — A Complete Market Guide (2026)
Data as of 10 October 2026. Uranium market structure is shown for 2024, the latest year every agency has counted, from the OECD-NEA/IAEA Uranium 2026: Resources, Production and Demand (the “Red Book”), with the Red Book’s expected 2025 output, the World Nuclear Association (WNA), the IAEA’s Nuclear Power Reactors in the World (2026 edition) and the producers’ 2025 filings.
Uranium is the fuel of nuclear power, and almost nothing else: in 2024 the world’s 441 commercial reactors needed 64,584 tonnes of uranium, and mines supplied 61,924 tonnes — 96% of it, with stockpiles and recycled material covering the rest. Three countries, Kazakhstan, Canada and Namibia, mined almost three-quarters of that, and one, Russia, owns 43% of the world’s capacity to enrich it into reactor fuel. The spot price averaged $85.14 a pound in 2024 and $85.34 over the 12 months to September 2026, ten times its level of 2000. This report is the free, big-picture primer on how the uranium market works: where uranium is mined, how the fuel cycle turns it into reactor fuel, how its price is set and which forces drive it. For the company-level data behind the story — every producer screened by production and resources — go to Metal Pilot’s uranium mining stock screener.
TL;DR & Key Takeaways
- What it is: a weakly radioactive metal valued for one isotope, uranium-235, whose fission heats a reactor’s water. Demand comes from the reactor fleet, which must be fuelled whatever the price.
- Who supplies it: mines produced 61,924 tU in 2024 (Red Book count); Kazakhstan alone mined 38%, and with Canada and Namibia 73%. Government-owned companies own 72% of output.
- Who uses it: the United States needs a third of the world’s uranium (21,500 tU) but mines almost none; China, at 16%, builds half the world’s new reactors.
- The balance: a structural primary gap — mines have covered 74–98% of reactor needs since 2015 (WNA count); inventories, ex-weapons material and underfeeding fill the rest. The choke point is downstream: Russia holds 43% of enrichment capacity.
- How it is priced: the uranium market trades privately, under long-term contracts; thin spot trades are assessed by two price-reporting agencies, UxC and TradeTech. Spot averaged $73.54/lb in 2025 and $85.34/lb over the 12 months to September 2026.
- The regime that favours it: none — uranium ignores the business cycle (correlation with the S&P 500 on monthly changes, 2000–2025: +0.10). It rises when reactor build-out meets producer discipline or fuel-cycle stress and collapses after accidents; miners amplify both.
Numbers to remember: mine production 61,924 tU (2024) · market size about $13.7 bn (2024) · spot $85.3/lb (12 months to Sep 2026); $73.54/lb (2025 average) · top producer Kazakhstan 23,270 tU · top consumer the United States 21,500 tU · mine output = 96% of reactor needs · Russia 43% of enrichment capacity · resource life ≈ 95 years.
Figure 1. Uranium at a glance, 2024 (price: 12 months to Sep 2026)
Notes to Figure 1
- Mine production — Table 4 · Filed | Red Book 2026 | Table 1.18 | “Historical uranium production”, row “Total”, column 2024 | 61,924 tU | p.78. The WNA counts 60,213 tU for the same year (Table 8).
- Market size — Derived | 61,924 tU × 2,599.8 lb U₃O₈ per tU × $85.14/lb (Table 3, 2024 average spot) = $13.7 bn; the value of one year’s mine output, not of any company. At the 2025 average and the Red Book’s expected 2025 output it is $12.6 bn.
- Spot price, 12 months to Sep 2026 — Market | Cameco | “Uranium Spot Price”, month-end industry average | Oct 2025–Sep 2026; the mean of the twelve month-end prices, computed here. For reference only: the 2025 calendar-year average was $73.54/lb (Table 3).
- Top producer — Table 4 · Filed | Red Book 2026 | Table 1.18 | “Kazakhstan”, 2024 | 23,270 tU | p.78.
- Top consumer — Table 6 · Filed | Red Book 2026 | Table 2.1 | “2024 uranium requirements (tU)", “United States” | 21,500 | p.106.
- Mine output ÷ reactor needs — Table 10 · Filed | Red Book 2026 | Primary sources of uranium supply | “world uranium production (61 924 tU) represented around 96% of 2024 world reactor requirements (64 584 tU)" | p.127. On the WNA’s count the 2024 figure is 90%; uranium is barely recycled, so this coverage ratio stands in for the “share of supply from recycling” a metal guide would show.
- Russia’s share of enrichment — Table 15 · Derived | Rosatom 29,133 ÷ world 68,076 thousand SWU/yr, 2025 | WNA, Uranium Enrichment, “World enrichment capacity” = 42.8%.
- Resource life — Table 9 · Derived | identified recoverable resources below $130/kgU, 5,864,500 tU (Red Book 2026, Table 1.2a, p.25) ÷ 2024 production 61,924 tU = 94.7 years, shown as 95. Resources are an estimate, not a reserve.
Figure data: Tables 3, 4, 6, 9, 10 and 15 — the OECD-NEA/IAEA Uranium 2026: Resources, Production and Demand (printed page numbers), Cameco’s industry-average uranium prices and the WNA’s Uranium Enrichment page; each tile’s citation is in the notes above.
How to read this guide: new to uranium? Read straight through from Section 1 — the fuel cycle is the key that unlocks everything else. Here for the investment case? Jump to Section 5 — how to own it, how to screen the miners and what actually drives this idiosyncratic market. Every technical term is defined in the Terms list at the end.
1. Uranium & the market basics
1.1 What uranium is — physical basics & quality
Uranium (chemical symbol U) is a dense, weakly radioactive metal whose economic value rests on one property: its isotope uranium-235 (U-235) can sustain a nuclear chain reaction, splitting atoms and releasing heat that boils water to spin a turbine. Natural uranium is only 0.7% U-235; the rest is U-238, which does not readily split. Most reactors therefore need the U-235 share raised to 3–5%, the step called enrichment. Almost all uranium demand is fuel for power reactors; research reactors, naval propulsion and medical isotopes take a small remainder. There is no monetary, jewellery or broad industrial pillar: uranium is a pure energy fuel.
A few terms define uranium’s quality and form, and they recur throughout this report:
- Grade (% U₃O₈) — the concentration of uranium oxide in the ore. The range is extreme: Cameco’s Cigar Lake reserves grade 16.33% U₃O₈, McArthur River 6.48%, while the Inkai in-situ leach deposits in Kazakhstan grade about 0.03%. Grade, more than the price, decides what is economic.
- Conventional vs. in-situ leach (ISL) — the two mining routes. Conventional mining digs ore from an open pit or underground and processes it in a mill. In-situ leach (also called in-situ recovery, ISR) pumps a leaching solution through a permeable sandstone orebody and recovers the dissolved uranium at surface; nothing is dug. ISL supplied 55% of world output in 2024, up from 16% in 2000.
- Yellowcake (U₃O₈) — the uranium oxide concentrate a mill or ISL plant produces, the form in which uranium is bought, sold and priced.
- Enrichment and SWU — the effort of raising the U-235 share, measured in separative work units (SWU) and priced separately.
- Primary vs. by-product — uranium mined for its own sake versus uranium recovered alongside another metal, as at BHP’s Olympic Dam copper mine in Australia.
The value chain — from ore to reactor. Uranium follows a longer path than any metal: mining and milling (yellowcake) → conversion (to uranium hexafluoride, UF₆, a gas) → enrichment (to 3–5% U-235) → fuel fabrication (uranium dioxide pellets in fuel assemblies) → the reactor → spent fuel (stored or reprocessed). A mine’s job ends at U₃O₈; each later step is a separate industry with its own, far smaller, set of suppliers. The strategic choke point is enrichment (Section 2.6).
Figure 2. The nuclear fuel cycle, ore to reactor
Source: this guide’s summary of the WNA’s Conversion and Deconversion and Uranium Enrichment pages; a conceptual diagram with no values.
From orebody to yellowcake. On the conventional route, ore is crushed and ground in a mill, and the uranium is dissolved out with acid or an alkaline solution, purified by ion exchange and precipitated as U₃O₈. ISL runs the same chemistry underground and needs no pit, waste rock or mill tailings, which is why low-grade sandstone deposits in Kazakhstan, Uzbekistan and the United States are mined that way.
1.2 Units & measurement conventions
This report uses the nuclear-industry convention, stated here so every later number is unambiguous. Quantities of uranium appear in two units the reader must convert between: tonnes of uranium metal (tU), used by the Red Book, the WNA and the IAEA for production, resources and reactor needs; and pounds of U₃O₈ (lb U₃O₈), the yellowcake unit in which the metal is priced. U₃O₈ is 84.8% uranium by mass, so 1 tU = 1.1792 t U₃O₈ = 2,599.8 lb U₃O₈, and a price of $1/lb U₃O₈ equals $2.60/kgU. Prices are in US dollars per pound of U₃O₈ ($/lb); enrichment is priced per SWU. Large quantities use M = million (Mlb, Mt).
Uranium figures split into flow and stock, and the two must not be confused:
- Flow — quantities per year: mine production (61,924 tU in 2024), reactor requirements (64,584 tU), the shortfall met by secondary supply.
- Stock — a level at a point in time: identified resources in the ground (5.86 m tU below $130/kgU at 1 January 2025), company reserves, and the inventories of natural and enriched uranium that utilities and governments hold.
Table 1. Uranium units and conversions
| Unit | Meaning | Typical magnitude in uranium | Conversion |
|---|---|---|---|
| tU | Tonne of uranium metal | World mine output ~62,000 tU/yr | = 2,599.8 lb U₃O₈ |
| lb U₃O₈ | Pound of yellowcake | The price-quote unit | = 0.000385 tU |
| Mlb U₃O₈ | Million pounds of yellowcake | A large mine: 15–20 Mlb/yr | = 384.6 tU |
| $/lb U₃O₈ | Price per pound | 2025 spot average $73.54 | × 2.6 = $/kgU |
| SWU | Separative work unit | US utilities bought 12.7 m SWU in 2025 | Priced separately ($/SWU) |
| % U₃O₈ | Ore grade | Cigar Lake 16.33%; Inkai 0.03% | 1% U₃O₈ = 0.848% U |
| Mt U | Million tonnes of uranium | Identified resources: 5.86 Mt U below $130/kgU | = 1,000,000 tU |
Source: OECD-NEA/IAEA, Uranium 2026: Resources, Production and Demand, Appendix 3 “Units” — “1% U₃O₈ = 0.848% U” and “1 USD/lb U₃O₈ = USD 2.6/kg U”, p.539; the tU-to-pound factor is derived here from the 0.848 ratio and 2,204.62 lb per tonne. Magnitudes from Tables 3, 4 and 9, Cameco’s 2025 annual report (Cigar Lake output and grades, Table 5) and the US EIA’s 2025 Uranium Marketing Annual Report, Table 16 (SWU), p.45.
Numbers intuition: the world’s biggest uranium mine, Cigar Lake, produced 19.1 Mlb (about 7,350 tU) in 2025; one large reactor of 1,000 MWe needs about 160 tU a year (the world fleet averages 162 tU per GWe of capacity). A year’s mine output is worth about $13 bn at recent prices — tiny next to copper or gold, yet it fuels the reactors that generated 2,845 TWh of electricity in 2025.
1.3 Pricing & benchmarks
Uranium has the most unusual pricing of any commodity in this series. There is no liquid exchange: buyers and sellers negotiate privately, and two price-reporting agencies, UxC and TradeTech, publish the reference prices. Two of them matter. The spot price covers one-off deliveries within a year; the Red Book puts spot at 15–25% of annual transactions, so it is thin, volatile and headline-grabbing. The long-term price is the reference for multi-year contracts, where most uranium changes hands: a utility running a reactor for decades values secure supply over a cheap cargo. The two can diverge widely, and long-term contracts often carry formulas that reference spot at delivery. Conversion and enrichment are bought separately, so the delivered cost of fuel blends several prices.
Agencies publish what buyers actually paid, and those averages lag the market. US reactor operators paid a weighted average of $58.46/lb for deliveries in 2025, and the Euratom Supply Agency (ESA) put EU utilities’ average spot purchase at $70.33/lb — against an industry spot average of $73.54. Since 2021 a physical trust has bought pounds outright on the spot market (Section 5.1), and CME lists a cash-settled futures contract on UxC’s spot price, but uranium remains a contract market at heart.
Table 2. Key uranium price benchmarks
| Benchmark | What it prices | Who sets it | Role | Source |
|---|---|---|---|---|
| Spot U₃O₈ | Delivery within a year | UxC, TradeTech (Cameco averages the two) | Headline price; thin, volatile | Market | Cameco | “Uranium Spot Price”, month-end |
| Long-term U₃O₈ | Multi-year contracts | UxC, TradeTech | Where most volume is priced | Market | Cameco | “Long-term Uranium Price”, month-end |
| UxC U₃O₈ futures | 250 lb, cash-settled on UxC spot | CME Group | Hedging tool; thin | Filed | CME contract specs | “Contract Unit” 250 pounds; “Financially Settled” |
| ESA spot and MAC-3 | Prices EU utilities paid | Euratom Supply Agency | Lagging, what was actually paid | Filed | ESA AR 2025 | Average prices | “ESA spot U₃O₈ price” | p.11 |
| US weighted-average price | Prices US operators paid | US EIA | Lagging; covers all contract types | Filed | EIA UMAR 2025 | Table 3 | “Total Purchases” | p.21 |
| Enrichment (SWU) | Separative work | Enrichers, by contract | Largest fuel-cycle cost after U₃O₈ | Filed | EIA UMAR 2025 | Table 16 | “Average price (US$ per SWU)" 108.70 | p.45 |
Source: Cameco, Uranium Price (month-end prices published by UxC and TradeTech, averaged by Cameco); CME Group, UxC Uranium U₃O₈ Futures Contract Specs; Euratom Supply Agency, Annual Report 2025; US EIA, 2025 Uranium Marketing Annual Report. The Role column is this guide’s summary.
The long-run price story is one of two great cycles. A 2003–2007 boom carried the annual spot average from $11.55 to $99.33/lb, with a month-end peak of $136 in June 2007. The bust that followed was compounded by the March 2011 Fukushima accident: spot ground down to $18.00 at the end of November 2016, and the annual average bottomed at $21.66 in 2017. Producer cuts, financial buying and the move away from Russian fuel then drove a recovery to $85.14 in 2024 — the long-term price followed more slowly and, at $81.96, stood above spot in 2025. Adjusted for US inflation, the 2007 average was worth $154 in 2025 dollars, double today’s level. Over 2016–2025 spot averaged $43.38 and the long-term price $47.32; over 2021–2025, $61.26 and $61.12.
Table 3. Average annual uranium price per pound of U₃O₈, 2000–2025
| Year | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Spot | $8.18/lb | $8.82/lb | $9.88/lb | $11.55/lb | $18.60/lb | $28.67/lb | $49.61/lb | $99.33/lb | $61.71/lb | $46.06/lb | $46.84/lb | $56.37/lb | $48.40/lb |
| Long-term | $9.45/lb | $10.12/lb | $10.62/lb | $12.10/lb | $20.35/lb | $30.66/lb | $49.90/lb | $90.83/lb | $82.50/lb | $65.50/lb | $60.92/lb | $66.79/lb | $60.12/lb |
| Spot, in 2025 dollars | $15.29/lb | $16.04/lb | $17.69/lb | $20.22/lb | $31.71/lb | $47.27/lb | $79.24/lb | $154.24/lb | $92.30/lb | $69.12/lb | $69.15/lb | $80.69/lb | $67.88/lb |
| Year | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Spot | $38.17/lb | $33.21/lb | $36.55/lb | $25.64/lb | $21.66/lb | $24.59/lb | $25.64/lb | $29.96/lb | $35.28/lb | $49.81/lb | $62.51/lb | $85.14/lb | $73.54/lb |
| Long-term | $54.12/lb | $46.46/lb | $46.29/lb | $39.00/lb | $31.89/lb | $30.38/lb | $31.75/lb | $34.62/lb | $36.81/lb | $49.75/lb | $58.20/lb | $78.88/lb | $81.96/lb |
| Spot, in 2025 dollars | $52.75/lb | $45.17/lb | $49.65/lb | $34.40/lb | $28.45/lb | $31.53/lb | $32.29/lb | $37.27/lb | $41.92/lb | $54.81/lb | $66.05/lb | $87.38/lb | $73.54/lb |
Notes to Table 3
- Cameco’s own 2025 annual report prints the same 2025 and 2024 averages — “Average spot price (US$/lb)" 73.54 and 85.14; “Average long-term price (US$/lb)" 81.96 and 78.88 (p.57). Long-term prices before May 2004 are TradeTech’s alone, per Cameco’s page.
- The month-end extremes quoted in the text are from the same series: $7.10 (December 2000), $136.00 (June 2007), $18.00 (November 2016) and $100.25 (January 2024); $69.63 (February 2011) is the last month-end before the Fukushima accident.
Source: Cameco, Uranium Price, series “Uranium Spot Price” and “Long-term Uranium Price” — month-end industry averages Cameco computes from UxC and TradeTech, 2000–2025; each annual value is the mean of the year’s twelve month-end prices, computed here. The real row is nominal spot × (US CPI-U 2025 ÷ CPI-U of the year), from FRED series CPIAUCSL (annual mean of monthly values; 2025 uses 11 months, October 2025 being unpublished). Multi-year averages in the text are computed here from the nominal rows.
Figure 3. Uranium’s two great cycles, 2000–2025: spot and long-term price
Figure data: Table 3, nominal rows. The long-term price swings less than spot and leads it at turning points, because contracts are signed ahead of need.
Current situation (2026, unresolved — for reference only): spot has averaged $87.41/lb from January to September 2026 and the long-term price $93.33, above any calendar-year average in the series; the September month-end figures were $89.63 and $96.50 — the long-term one the highest month-end in the series, above the $95.00 plateau of 2007–08 (Cameco series, Figure 1, note 3). These are prices of a year still in progress and anchor none of the analysis below.
2. Supply, demand & the market balance
2.1 Where uranium is mined — deposits & geology
Uranium supply is geographically concentrated, and the deposit type dictates the mining method and the cost. The richest deposits on Earth are the unconformity orebodies of Canada’s Athabasca Basin in Saskatchewan — McArthur River and Cigar Lake — mined underground at reserve grades of 6–16% U₃O₈ (Table 5, note 4). The largest volume comes from sandstone deposits leached in place: Kazakhstan, Uzbekistan and parts of the United States and Australia. Namibia’s Husab and Rössing are large, low-grade open pits; Australia’s Olympic Dam yields uranium as a by-product of a giant copper mine; Niger’s sandstone mines at Arlit were a steady source until the 2023 coup and the 2025 nationalisation of the Somaïr mine; the Red Book expects no Nigerien output in 2025.
In 2024 Kazakhstan, Canada, Namibia, Uzbekistan and Australia produced 88.5% of the world’s uranium — Kazakhstan alone 37.6%. Eleven of the 18 producing countries mine only through government-owned companies, and government-owned companies owned 72% of all 2024 output.
Table 4. Leading uranium-mining countries, 2022–2025
| Rank | Country | 2022 | 2023 | 2024 | Share of world, 2024 | 2025e |
|---|---|---|---|---|---|---|
| 1 | Kazakhstan | 21,279 tU | 21,112 tU | 23,270 tU | 37.6% | 25,789 tU |
| 2 | Canada | 7,380 tU | 10,986 tU | 14,327 tU | 23.1% | 14,251 tU |
| 3 | Namibia 1 | 5,612 tU | 6,986 tU | 7,332 tU | 11.8% | 8,720 tU |
| 4 | Uzbekistan | 3,561 tU | 4,050 tU | 5,200 tU | 8.4% | 6,000 tU |
| 5 | Australia | 4,555 tU | 4,686 tU | 4,656 tU | 7.5% | 5,395 tU |
| 6 | Russia | 2,508 tU | 2,710 tU | 2,738 tU | 4.4% | 2,620 tU |
| 7 | China | 1,550 tU | 1,728 tU | 1,956 tU | 3.2% | 2,160 tU |
| 8 | Niger 1, 2 | 2,020 tU | 1,130 tU | 959 tU | 1.5% | 0 tU |
| 9 | India 1 | 488 tU | 485 tU | 485 tU | 0.8% | 485 tU |
| 10 | Ukraine | 120 tU | 350 tU | 329 tU | 0.5% | 300 tU |
| — | Rest of world | 417 tU | 340 tU | 672 tU | 1.1% | 354 tU |
| — | World total | 49,490 tU | 54,563 tU | 61,924 tU | 100% | 66,074 tU |
Notes to Table 4
- Secretariat estimates in the Red Book.
- Niger’s 2025 figure is the Red Book’s expected zero: in July 2025 the government announced the nationalisation of the Orano-majority Somaïr (Arlit) mine (Table 1.17, p.77).
- “Rest of world” is the world total less the ten rows; in 2024 it is mainly the United States (260 tU) and South Africa (235 tU). The Red Book gives no expected 2025 value for the United States (“NA”), so the 2025 world total excludes it.
Source: OECD-NEA/IAEA, Uranium 2026: Resources, Production and Demand, Table 1.18 “Historical uranium production”, columns 2022, 2023, 2024 and “2025 (expected)", pp.77–78; the government-owned share from Table 1.19 “Ownership of uranium production” (44,575 tU, 72.0%), p.89. Shares and the rest-of-world row are computed here. The Red Book’s count is a little above the WNA’s (61,924 against 60,213 tU in 2024), mainly on Uzbekistan and China; this guide keeps the two apart (Section 8.1).
Figure 4. Leading uranium-mining countries, 2024 (tU)
Figure data: Table 4, column 2024.
At the asset level uranium is the most concentrated market in this series. The two Athabasca mines produced about 13,150 tU in 2025 between them — roughly a fifth of world output — and the eight mines in Table 5 about 31,450 tU — close to half of the Red Book’s expected 2025 world output. The grade spread is the industry’s defining contrast: Cigar Lake ore is about five hundred times richer than Inkai’s, yet the Kazakh ISL wellfields are among the cheapest producers on the cost curve (Section 5.2).
Table 5. Largest uranium mines, 2025
| Mine | Country | Operator and owners | Output, 2025 | Method | Source |
|---|---|---|---|---|---|
| Cigar Lake | Canada | Cameco 54.547% (operator), Orano 40.453%, others 5.0% | 7,347 tU | Underground; 16.33% U₃O₈ reserves | Filed | CCJ AR 2025 | Cigar Lake | “2025 production 10.4 million pounds (19.1 million pounds on 100% basis)" | p.82 Derived | ÷ 2,599.8 |
| McArthur River / Key Lake | Canada | Cameco 69.805% (operator), Orano 30.195% | 5,808 tU | Underground mine, separate mill | Filed | CCJ AR 2025 | McArthur River/Key Lake | “2025 production 10.5 million pounds (15.1 million pounds on 100% basis)" | p.78 Derived | ÷ 2,599.8 |
| KATCO (Moinkum, Tortkuduk) | Kazakhstan | Orano (operator), Kazatomprom 49% 1 | 3,714 tU | ISL | Filed | KAP IAR 2025 | Uranium segment production by operation | “JV KATCO LLP” 3,714 | p.36 |
| Karatau (Budenovskoye 2) | Kazakhstan | Kazatomprom 50%, Uranium One 50% | 3,622 tU | ISL | Filed | KAP IAR 2025 | production by operation | “Karatau LLP” 3,622 | p.36 |
| Inkai | Kazakhstan | Kazatomprom 60%, Cameco 40% | 3,215 tU | ISL; 0.03% U₃O₈ reserves | Filed | KAP IAR 2025 | production by operation | “JV Inkai LLP” 3,215 | p.36 |
| South Mining Chemical Co. (Akdala, Inkai 4) | Kazakhstan | Kazatomprom 30%, Uranium One | 2,845 tU | ISL | Filed | KAP IAR 2025 | production by operation | “JV South Mining Chemical Company LLP” 2,845 | p.36 |
| Olympic Dam | Australia | BHP 100% | 2,675 tU 2 | By-product of copper; underground | Filed | BHP AR 2025 | Production | “Payable metal in concentrate (t)" uranium, 3,154 | p.201 Derived | × 0.848 |
| Akbastau (Budenovskoye 1, 3, 4) | Kazakhstan | Kazatomprom 50%, Uranium One | 2,227 tU | ISL | Filed | KAP IAR 2025 | production by operation | “JV Akbastau JSC” 2,227 | p.36 |
Notes to Table 5
Output is each mine’s whole output, at 100%, whatever the owners’ shares.
- Kazatomprom gives its KATCO stake as 49%; Orano’s 2025 report prints both 51%/49% and 49%/51% for the same joint venture (pp.34 and 234), so only Kazatomprom’s figure is shown.
- BHP’s year ends on 30 June: the figure is the year to June 2025, reported as uranium oxide concentrate and converted here at 0.848 tU per tonne of U₃O₈.
- Namibia’s Husab, run by CGN’s Swakop Uranium, produced 4,437 tU in 2024 (Red Book 2026, Table 1.17, p.77) and would rank third, but its 2025 output is not in the documents checked, so it is not ranked; nor is Rössing.
- Grades are those of Cameco’s reserve statement at 31 December 2025: “Cigar Lake UG” 16.33% and “Inkai ISR” 0.03% U₃O₈ (Cameco AR 2025, Mineral reserves, p.110); McArthur River’s is 6.48%.
Source: the operators’ filings, row by row as cited — Cameco 2025 Annual Report, Kazatomprom 2025 Integrated Annual Report (annual reports → 2025), BHP Annual Report 2025 and, for KATCO’s operator, Orano’s Annual Activity Report 2025. Pounds of U₃O₈ are converted at 2,599.8 lb per tU (Table 1). Output is the whole mine (100%), not the owners’ shares. Screen every producing asset on Metal Pilot’s uranium mining project screener.
2.2 Demand & consumption
Uranium demand is the reactor fleet. At 1 January 2025 the Red Book counted 441 commercial reactors in 31 countries (including 23 in suspended operation, mostly in Japan) with 398 GWe of capacity, and put their 2024 needs at 64,584 tU — about 146 tU per reactor, or 162 tU per GWe. That demand is highly price-inelastic: fuel is a small part of what it costs to run a reactor, a reactor must be refuelled whatever the uranium price, and utilities buy years ahead under contract.
The fleet is concentrated. The United States, with 94 reactors, needed 21,500 tU in 2024 — a third of the world total — followed by China (10,500 tU), France (7,200), South Korea (4,097) and Russia (4,000). The trend is set by China: of the 72 reactors under construction at the end of 2025, 36 were Chinese.
Table 6. Uranium requirements by country, 2024
| Country | Reactors operational | Uranium required, 2024 | Share of world |
|---|---|---|---|
| United States | 94 | 21,500 tU | 33.3% |
| China 1 | 58 | 10,500 tU | 16.3% |
| France | 57 | 7,200 tU | 11.1% |
| South Korea | 26 | 4,097 tU | 6.3% |
| Russia | 36 | 4,000 tU | 6.2% |
| Japan 2 | 33 | 2,021 tU | 3.1% |
| India | 24 | 1,650 tU | 2.6% |
| Canada | 17 | 1,512 tU | 2.3% |
| Ukraine | 15 | 1,500 tU | 2.3% |
| Spain | 7 | 1,208 tU | 1.9% |
| Sweden | 6 | 1,121 tU | 1.7% |
| Rest of world | 68 | 8,275 tU | 12.8% |
| World total | 441 | 64,584 tU | 100% |
Notes to Table 6
Reactors are counted at 1 January 2025.
- Excludes Chinese Taipei (one reactor, 150 tU), which the Red Book counts in the world total only.
- Japan’s 33 reactors include those in suspended operation; at the end of 2025 the IAEA counted 14 operating and 19 suspended.
Source: OECD-NEA/IAEA, Uranium 2026: Resources, Production and Demand, Table 2.1 “Nuclear power plant data summary”, columns “Operational reactors” and “2024 uranium requirements (tU)", p.106; reactors under construction from the IAEA’s Nuclear Power Reactors in the World, 2026 edition, Table 1, pp.8–9. Shares and the rest-of-world row are computed here. Requirements exclude MOX fuel.
Figure 5. Uranium requirements by country, 2024
Figure data: Table 6 — the five largest consumers, with every smaller fleet (Japan, India, Canada and the rest) grouped as “Rest of world”.
No agency publishes an annual series of uranium requirements by country back to 2000, but reactor output is the same story told in electricity: requirements track generation, at roughly 24 tU per TWh. World nuclear generation fell 10.7% between 2010 and 2012 as Japan idled its fleet after Fukushima and Germany began its exit, then recovered to a record 2,845 TWh in 2025. Under that flat total sits a geographic shift: China’s output grew 29-fold after 2000, to 17% of the world in 2025, while Europe’s fell 28%.
Table 7. Nuclear electricity generation by region, 2000–2025
| Region | 2000 | 2005 | 2010 | 2015 | 2020 | 2024 | 2025 |
|---|---|---|---|---|---|---|---|
| North America | 874 TWh | 925 TWh | 946 TWh | 952 TWh | 940 TWh | 920 TWh | 921 TWh |
| Europe 1 | 1,049 TWh | 1,110 TWh | 1,032 TWh | 968 TWh | 833 TWh | 766 TWh | 759 TWh |
| China | 17 TWh | 53 TWh | 75 TWh | 171 TWh | 366 TWh | 451 TWh | 485 TWh |
| Asia Pacific excl. China | 483 TWh | 500 TWh | 508 TWh | 248 TWh | 289 TWh | 363 TWh | 359 TWh |
| CIS | 133 TWh | 152 TWh | 173 TWh | 198 TWh | 219 TWh | 234 TWh | 239 TWh |
| Middle East | 0 TWh | 0 TWh | 0 TWh | 4 TWh | 8 TWh | 46 TWh | 44 TWh |
| South & Central America | 12 TWh | 17 TWh | 22 TWh | 22 TWh | 24 TWh | 26 TWh | 27 TWh |
| Africa | 13 TWh | 11 TWh | 14 TWh | 12 TWh | 12 TWh | 8 TWh | 11 TWh |
| World | 2,581 TWh | 2,769 TWh | 2,769 TWh | 2,576 TWh | 2,691 TWh | 2,815 TWh | 2,845 TWh |
Notes to Table 7
- EI counts Ukraine in Europe, not in the CIS.
Source: Energy Institute, Statistical Review of World Energy 2026 — data workbook, sheet “Nuclear Generation - TWh”, rows “Total North America”, “Total Europe”, “China”, “Total Asia Pacific”, “Total CIS”, “Total Middle East”, “Total S. & Cent. America”, “Total Africa” and “Total World”, 2000–2025 in five-year steps plus 2024; the full annual series feeds Figure 6. China is split out from Asia Pacific given its build-out (Asia Pacific excl. China = the regional total less China, computed here); the other regions follow the EI grouping. Rounded to the nearest TWh. The ratio of about 24 tU per TWh is this guide’s division of 2024 requirements (64,584 tU, Table 6) by the Red Book’s 2024 generation (2,727 TWh, Table 2.2, p.107).
Figure 6. Nuclear generation by region, 2000–2025 (TWh)
Figure data: Table 7, annual values; the gray band is South & Central America, the Middle East and Africa together.
2.3 Supply: producing countries
Mine output has been volatile. On the WNA’s count it peaked at 63,207 tU in 2016, fell to 47,731 tU in 2020 as producers idled mines into a glut and then a pandemic, and recovered to 60,213 tU in 2024 as McArthur River restarted and Kazakhstan lifted output. The Red Book expects 66,074 tU for 2025 on its own, slightly higher, count. Supply is also concentrated by ownership: Kazatomprom, Rosatom’s Uranium One, Orano, CGN, CNNC and Navoi Mining are all state-controlled (Table 21).
Table 8. World uranium mine production, 2015–2025
| Year | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| WNA count | 60,342 tU | 63,207 tU | 60,462 tU | 54,154 tU | 54,742 tU | 47,731 tU | 47,805 tU | 49,614 tU | 54,433 tU | 60,213 tU | n/p |
| Red Book count | n/p | n/p | n/p | n/p | n/p | n/p | n/p | 49,490 tU | 54,563 tU | 61,924 tU | 66,074 tU e |
Notes to Table 8
- The 2025 Red Book value is the expected figure (“e”); the WNA has not yet published 2025. “n/p” = not published in the edition cited.
Source: WNA, World Uranium Mining Production, table “Production from mines (tonnes U)”, row “World total tU”, 2015–2024; OECD-NEA/IAEA, Uranium 2026, Table 1.18, row “Total”, 2022–2025 expected, p.78. The two counts differ by up to 2.8%, mostly on Uzbekistan and China; the WNA row is the 2015–2024 series used for the balance (Table 10).
Figure 7. World uranium mine production, 2015–2024 (tU)
Figure data: Table 8, WNA row.
Uranium is not scarce in the ground. The Red Book’s identified recoverable resources — deposits delineated well enough to estimate, at a cost below $130 per kgU — were 5.86 m tU at 1 January 2025, about 95 years of 2024 output; below $260/kgU they reach 8.1 m tU. Australia holds 29%, mostly at Olympic Dam, followed by Kazakhstan, Canada and Russia. Resources are estimates on each country’s own basis, and much of Australia’s sits in a by-product orebody that will be mined only as fast as its copper.
Table 9. Identified recoverable uranium resources by country, below $130/kgU, 2025
| Rank | Country | Resources | Share of world |
|---|---|---|---|
| 1 | Australia | 1,701,600 tU | 29.0% |
| 2 | Kazakhstan | 827,500 tU | 14.1% |
| 3 | Canada | 551,700 tU | 9.4% |
| 4 | Russia | 471,700 tU | 8.0% |
| 5 | Namibia 1 | 392,800 tU | 6.7% |
| 6 | Niger 1 | 351,100 tU | 6.0% |
| 7 | China | 325,400 tU | 5.5% |
| 8 | South Africa 1 | 320,900 tU | 5.5% |
| 9 | Brazil | 167,800 tU | 2.9% |
| 10 | Ukraine | 161,300 tU | 2.8% |
| — | Rest of world | 592,700 tU | 10.1% |
| — | World total | 5,864,500 tU | 100% |
Notes to Table 9
- Secretariat estimates in the Red Book.
- The WNA’s World Uranium Mining Production page prints a “to $130/kg U” world total of 7,824,600 tU, which does not match the Red Book it cites; this table uses the Red Book directly.
Source: OECD-NEA/IAEA, Uranium 2026, Table 1.2a “Identified recoverable uranium resources”, column "<USD 130/kgU”, as of 1 January 2025, pp.24–25; the total below $260/kgU is 8,104,500 tU. Resources are reasonably assured plus inferred, recoverable, rounded to the nearest 100 tU. Shares, the rest-of-world row and the resource life are computed here.
2.4 The supply–demand balance
Uranium’s balance has one defining feature: mines have not covered reactor needs in any year since about 1990, when the Red Book dates the end of mine overproduction. The gap is filled by secondary supply — utility, producer and government inventories, re-enriched depleted uranium, recycled material and, until 2013, Russian weapons uranium (Section 2.5). On the WNA’s count, mine output covered 98% of requirements in 2015, fell to 74% in 2020 after a decade of low prices and the pandemic shut mines, and recovered to 90% in 2024. The Red Book, counting both sides differently, puts 2024 coverage at 96%. Either way, the structural question is whether new mines arrive before the inventories that have bridged the gap run down — the outlook in Section 6.
Table 10. Mine production against reactor requirements, 2015–2024
| Year | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|---|---|
| Mine production | 60,342 tU | 63,207 tU | 60,462 tU | 54,154 tU | 54,742 tU | 47,731 tU | 47,805 tU | 49,614 tU | 54,433 tU | 60,213 tU |
| Mine output ÷ requirements | 98% | 96% | 93% | 80% | 81% | 74% | 76% | 76% | 83% | 90% |
| Requirements, implied | ~61,600 tU | ~65,800 tU | ~65,000 tU | ~67,700 tU | ~67,600 tU | ~64,500 tU | ~62,900 tU | ~65,300 tU | ~65,600 tU | ~66,900 tU |
| Shortfall met by secondary supply | ~1,300 tU | ~2,600 tU | ~4,500 tU | ~13,500 tU | ~12,900 tU | ~16,800 tU | ~15,100 tU | ~15,700 tU | ~11,200 tU | ~6,700 tU |
Notes to Table 10
- The WNA prints production and the coverage ratio (its row ”% of world demand”), not requirements. The implied requirement is production ÷ coverage, rounded to the nearest 100 tU; because the coverage ratio is printed as a whole percentage, each implied value is uncertain by about ±400 tU, and the shortfall by the same.
- The Red Book’s own 2024 pair is 61,924 tU of production against 64,584 tU of requirements — a shortfall of 2,660 tU and coverage of 95.9% (p.127).
Source: WNA, World Uranium Mining Production, table “Production from mines (tonnes U)”, rows “World total tU” and ”% of world demand”, 2015–2024. The implied-requirement and shortfall rows are computed here, as note 1 explains. The Red Book comparison is OECD-NEA/IAEA, Uranium 2026, “Primary sources of uranium supply”, p.127.
Figure 8. Mine output as a share of reactor requirements, 2015–2024 (%)
Figure data: Table 10, row “Mine output ÷ requirements”; the dashed line marks 100%, the level at which mines alone would fuel the fleet.
At the country level the market splits into a handful of net suppliers that mine far more than they burn — Kazakhstan, Namibia, Uzbekistan and Australia run no power reactors at all — and the large net importers. The United States is the extreme case: it needs 21,500 tU a year and mined 260 tU in 2024. China, France, South Korea and Japan also depend on imports; Russia, despite its mines, consumes more than it produces.
Table 11. Uranium net positions, major countries, 2024
| Country | Mine production | Reactor requirements | Net position | Role |
|---|---|---|---|---|
| Kazakhstan | 23,270 tU | 0 tU | +23,270 tU | Net supplier |
| Canada | 14,327 tU | 1,512 tU | +12,815 tU | Net supplier |
| Namibia | 7,332 tU | 0 tU | +7,332 tU | Net supplier |
| Uzbekistan | 5,200 tU | 0 tU | +5,200 tU | Net supplier |
| Australia | 4,656 tU | 0 tU | +4,656 tU | Net supplier |
| India | 485 tU | 1,650 tU | −1,165 tU | Net importer |
| Ukraine | 329 tU | 1,500 tU | −1,171 tU | Net importer |
| Russia | 2,738 tU | 4,000 tU | −1,262 tU | Net importer |
| Japan | 0 tU | 2,021 tU | −2,021 tU | Net importer |
| South Korea | 0 tU | 4,097 tU | −4,097 tU | Net importer |
| France | 0 tU | 7,200 tU | −7,200 tU | Net importer |
| China | 1,956 tU | 10,500 tU | −8,544 tU | Net importer |
| United States | 260 tU | 21,500 tU | −21,240 tU | Net importer |
Source: OECD-NEA/IAEA, Uranium 2026 — production from Table 1.18, column 2024, pp.77–78; requirements from Table 2.1, column “2024 uranium requirements (tU)", p.106 (a country absent from Table 2.1 has no power reactor and is entered as 0). Net position = production − requirements, computed here; it is a physical balance, not a trade statistic, and ignores inventories and re-exports. The Role column is this guide’s label.
Figure 9. Uranium net positions, major countries, 2024 (tU)
Figure data: Table 11 — the five net suppliers and the four largest net importers; India, Ukraine, Russia and Japan, each with net imports below 2,100 tU, are left out.
Inventories are the cover. Uranium has no exchange stocks to watch; its shock absorber is the inventory utilities hold. At the end of 2025 EU utilities owned 42,522 tU of natural-uranium equivalent — 3.6 years of their average annual needs — and US reactor operators about 45,500 tU, roughly 2.1 years of US requirements. US brokers and traders held another 17,000 tU; the Red Book notes their holdings at the end of 2024 were more than six times the end-2015 level. These are working stocks spread across the fuel cycle, not idle metal; when utilities decide they hold too little, their restocking lands on a thin spot market, as it did after 2022, when EU utilities began stockpiling.
Table 12. Utility uranium inventories and the cover they represent
| Holder | Inventory, end-year | Year | Annual requirement | Cover | Source |
|---|---|---|---|---|---|
| EU utilities | 42,522 tU | 2025 | 11,835 tU | 3.6 years | Filed | ESA AR 2025 | Inventories | “inventories owned by EU utilities totalled 42 522 tU” | p.16; requirement "(11 835 tU per year)" | p.17 |
| EU utilities | 51,892 tU | 2015 | — | — | Filed | Red Book 2026 | Table 2.5 | “Inventories held by EU utilities” | p.131 |
| US reactor owners and operators | 45,487 tU | 2025 1 | 21,500 tU | 2.1 years | Filed | EIA UMAR 2025 | Table 23 | “Owners and operators of U.S. civilian nuclear power reactors” 118,256 thousand lb U₃O₈e | p.59 Derived | ÷ 2,599.8 |
| US brokers and traders | 17,154 tU | 2025 1 | — | — | Filed | EIA UMAR 2025 | Table 23 | “U.S. brokers and traders” 44,597 thousand lb | p.59 Derived | ÷ 2,599.8 |
| All US commercial holders | 65,384 tU | 2025 1 | — | — | Filed | EIA UMAR 2025 | Table 23 | “Total commercial inventories” 169,986 thousand lb | p.59 Derived | ÷ 2,599.8 |
Notes to Table 12
- Preliminary 2025 data in the EIA report.
- The US requirement is the Red Book’s 2024 figure for the United States (Table 6). Cover = inventory ÷ annual requirement, computed here. EU inventories exclude the UK from 2021; the Red Book’s 2015 figure includes it.
- The Red Book notes that US traders and brokers held about 13,430 tU at the end of 2024, more than six times their end-2015 level (p.131).
Source: Euratom Supply Agency, Annual Report 2025, section 1.5 “Inventories” (and, on stockpiling since 2022, “EU utilities have been stockpiling nuclear materials and fuel”, p.10); US EIA, 2025 Uranium Marketing Annual Report, Table 23 “Inventories of uranium by owner as of end of year”; OECD-NEA/IAEA, Uranium 2026, Table 2.5. US pounds of U₃O₈ equivalent converted at 2,599.8 lb per tU (Table 1). China publishes no inventory series; the Red Book cites a WNA estimate from import statistics of about 147,000 tU at 1 January 2025 (Uranium 2026, p.130), and nothing comparable exists for Russia.
2.5 Supply structure
Primary supply is all mined, and the mining method has shifted decisively. In-situ leach supplied 55.2% of 2024 production, underground mining 26.1% — boosted by McArthur River’s restart — open pits 13.8% and co-product and by-product recovery 4.7%, almost all of it Olympic Dam, where uranium rides with copper and responds to the copper price rather than its own. ISL lowers the industry’s cost base and needs no tailings dam, but its wellfields deplete within a few years, so Kazakh output depends on continual new drilling — and on sulfuric acid, whose shortages the Red Book lists among Kazakhstan’s recent supply disruptions.
The rest of supply is secondary: material mined years ago or made from something else. The Red Book lists stocks and inventories of natural and enriched uranium, civil and military; fuel from reprocessing spent fuel and surplus plutonium; underfeeding, where enrichers run extra separative work to use less natural uranium; and the re-enrichment of depleted uranium (“tails”). The largest historical source, the Megatons to Megawatts agreement, converted 500 tonnes of Russian warhead uranium into reactor fuel for the United States between 1993 and 2013. Recycling is small: mixed-oxide (MOX) fuel saved EU utilities an estimated 605 tU in 2024, and reprocessed uranium is under 1% of world requirements.
Table 13. Uranium supply structure, 2024
| Supply source | Share, 2024 | Trend | Price response | Source |
|---|---|---|---|---|
| Mine — in-situ leach | 55.2% of mine output | 59.8% in 2022; 57.7% in 2025 (prelim.) | Medium: low cost, but needs constant drilling | Filed | Red Book 2026 | Table 1.22 | “ISL” | p.92 |
| Mine — underground | 26.1% | 18.0% in 2022, after McArthur River’s restart | Low: long lead times | Filed | Red Book 2026 | Table 1.22 | “Underground mining” | p.92 |
| Mine — open pit | 13.8% | 15.9% in 2022 | Low | Filed | Red Book 2026 | Table 1.22 | “Open-pit mining” | p.92 |
| Mine — co-product and by-product | 4.7% | 6.1% in 2022 | Very low: follows copper | Filed | Red Book 2026 | Table 1.22 | “Co-product/by-product” | p.92 |
| Mine — heap leach and other | 0.2% | — | — | Filed | Red Book 2026 | Table 1.22 | “Heap leaching”, “Other” | p.92 |
| Secondary supply (gap to requirements) | ~4% of requirements (2,660 tU) | Down from ~26% in 2020 (WNA count) | Finite; falls as stocks are drawn | Derived | Table 10: requirements − production |
| MOX fuel in EU reactors | 605 tU saved | — | Policy-driven | Filed | Red Book 2026 | MOX | “reduced natural uranium requirements in the EU by an estimated 605 tU in 2024” | p.133 |
| Reprocessed uranium | < 1% of requirements | — | Policy-driven | Filed | Red Book 2026 | RepU | “less than 1% of projected annual world requirements” | p.133 |
Source: OECD-NEA/IAEA, Uranium 2026, Table 1.22 “World uranium production by production method (2022-2025)", p.92, and “Secondary sources of uranium supply”, pp.128–133; the ISL share of 16% in 2000 quoted in Section 1.1 is from the WNA’s In Situ Leach Mining of Uranium page. Mining-method shares are of mine output; secondary supply is a share of requirements, on the Red Book’s count (≈ 10% on the WNA’s, Table 10). The Price response column is this guide’s judgement.
Figure 10. Uranium production by mining method, 2024 (% of mine output)
Figure data: Table 13, mine rows.
2.6 Trade flows & the fuel-cycle choke point
Uranium’s trade map is really a map of the fuel cycle. It is compact and valuable — a year of world mine output weighs about 73,000 tonnes of U₃O₈ — so there are no bulk-shipping choke points of the oil or copper kind. Yellowcake moves from a few mining countries to conversion plants in Canada, France, the United States, Russia and China, then to enrichment, then to fuel fabricators near the reactors. Trade statistics for uranium are patchy — several producers report nothing to UN Comtrade — so the clearest picture comes from the buyers. In 2025 Canada and Kazakhstan supplied more than half of what both EU and US utilities bought; Russia supplied 16% of the EU’s natural uranium, while the US figure for Russian origin is withheld.
Table 14. Where EU and US utilities bought natural uranium, 2025
| Origin | EU utilities | Share of EU | US operators | Share of US |
|---|---|---|---|---|
| Canada | 5,384 tU | 36.7% | 15,207 klb U₃O₈e | 32.4% |
| Kazakhstan | 2,981 tU | 20.3% | 13,331 klb U₃O₈e | 28.4% |
| Russia | 2,346 tU | 16.0% | withheld | withheld |
| Australia | 1,379 tU | 9.4% | 7,077 klb U₃O₈e | 15.1% |
| Uzbekistan | 1,522 tU | 10.4% | 3,258 klb U₃O₈e | 7.0% |
| Namibia | 842 tU | 5.7% | 1,652 klb U₃O₈e | 3.5% |
| United States | not listed 2 | — | 3,376 klb U₃O₈e | 7.2% |
| Other and withheld 1 | 224 tU | 1.5% | 2,970 klb U₃O₈e | 6.3% |
| Total | 14,678 tU | 100% | 46,871 klb U₃O₈e | 100% |
Notes to Table 14
- EU: Niger 33, Malawi 28, South Africa 26, Kyrgyzstan 22 and 115 tU of underfeeding, mixed and unknown origin. US: Ukraine 1,056, Niger 213, Malawi 73 and every cell the EIA withholds to protect company data, Russia’s included; computed as the total less the rows shown.
- The ESA’s origin table has no United States row; any US-origin material delivered to EU utilities sits in “Other”.
Source: Euratom Supply Agency, Annual Report 2025, Table 3 “Origin of uranium delivered to EU utilities in 2025 (tU)", columns “Quantity” and “Share (%)", p.12; US EIA, 2025 Uranium Marketing Annual Report, Table 3 “Uranium purchased by owners and operators of U.S. civilian nuclear power reactors by origin country and delivery year”, column “Deliveries in 2025”, p.21. US shares are computed here. The two blocs report in different units, so the shares, not the quantities, are comparable; the US total equals about 18,000 tU.
Figure 11. Where EU and US utilities bought their uranium, 2025 (% of purchases)
Figure data: Table 14, share columns; the US Russia figure is withheld by the EIA, so its bar is empty.
The choke point is enrichment. Four operators hold almost all of the world’s 68 m SWU of enrichment capacity, and Rosatom alone has 43% — more than Urenco and Orano, the two Western enrichers, combined at 36%. Conversion is less concentrated but tight in the West. After Russia’s 2022 invasion of Ukraine, Western utilities began cutting Russian fuel services, and in August 2024 the United States banned imports of Russian low-enriched uranium, with Department of Energy waivers allowed until the start of 2028 and the ban itself running to the end of 2040. Russia still supplied 26% of the enrichment US operators bought in 2025 and 23% of the EU’s. Urenco and Orano are adding capacity — 700,000 SWU a year in New Mexico by 2027 and 2.5 m SWU at France’s Georges Besse II by 2030 — but the West will lean on Russian enrichment for years.
Table 15. The fuel-cycle choke points: conversion and enrichment capacity
| Stage | Operator | Country | Capacity | Share of world | Source |
|---|---|---|---|---|---|
| Enrichment, 2025 | Rosatom | Russia | 29,133 thousand SWU/yr | 42.8% | Filed | WNA enrichment page | “World enrichment capacity” | “Rosatom”, 2025 |
| Enrichment, 2025 | Urenco | UK, Germany, Netherlands, US | 17,300 thousand SWU/yr | 25.4% | Filed | WNA enrichment page | “Urenco”, 2025 |
| Enrichment, 2025 | CNNC 1 | China | 13,789 thousand SWU/yr | 20.3% | Filed | WNA enrichment page | “CNNC*", 2025 |
| Enrichment, 2025 | Orano | France | 7,500 thousand SWU/yr | 11.0% | Filed | WNA enrichment page | “Orano”, 2025 |
| Enrichment, 2025 | Others (INB, JNFL) | Brazil, Japan | 355 thousand SWU/yr | 0.5% | Filed | WNA enrichment page | “Other (INB, JNFL)", 2025 |
| Conversion, 2022 | CNNC 1 | China | 15,000 tU/yr | 24.2% | Filed | WNA conversion page | “Estimated world primary conversion capacity” | “CNNC (est.)*" |
| Conversion, 2022 | Orano | France | 15,000 tU/yr | 24.2% | Filed | WNA conversion page | “Orano” |
| Conversion, 2022 | Cameco | Canada | 12,500 tU/yr | 20.2% | Filed | WNA conversion page | “Cameco” |
| Conversion, 2022 | Rosatom | Russia | 12,500 tU/yr | 20.2% | Filed | WNA conversion page | “Rosatom (est.)" |
| Conversion, 2022 | ConverDyn | United States | 7,000 tU/yr | 11.3% | Filed | WNA conversion page | “ConverDyn” |
| Russian share of SWU bought, 2025 | — | United States | 3,284 of 12,710 thousand SWU | 25.8% | Filed | EIA UMAR 2025 | Table 16 | “Russia”, “Total” | p.45 |
| Russian share of SWU bought, 2025 | — | EU | 2,735 of 12,130 tSW | 22.6% | Filed | ESA AR 2025 | Table 5 | “Russia” 22.55 | p.15 |
Notes to Table 15
- The WNA estimates China’s enrichment and conversion capacity on the assumption that China builds enough to supply its own fleet.
- Conversion figures are licensed capacity in 2022, the latest year the WNA publishes; world capacity was 62,000 tU against 42,000 tU of production, and ConverDyn’s Metropolis plant produced nothing that year.
Source: WNA, Uranium Enrichment, table “World enrichment capacity” (from the WNA Nuclear Fuel Report 2025), columns 2025 and 2030; WNA, Conversion and Deconversion, table “Estimated world primary conversion capacity and UF₆ production 2022” (from the Nuclear Fuel Report 2023); US EIA, 2025 Uranium Marketing Annual Report, Table 16; Euratom Supply Agency, Annual Report 2025, Table 5. Shares are computed here. The US import ban, its waivers and the Urenco and Orano expansions are described on the same WNA enrichment page.
Figure 12. World uranium enrichment capacity by operator, 2025 (thousand SWU/yr)
Figure data: Table 15, enrichment rows. By 2030, on the WNA’s projection, Rosatom’s share falls to 38% as CNNC, Urenco and Orano expand.
2.7 Market organisations & supply coordination
Uranium has no OPEC and no quota system, but it is among the most state-dominated markets in this series, and coordination shows up in three places. First, state producers’ output decisions: Kazakhstan’s state holds 75% of Kazatomprom, the largest producer, whose production guidance is among the market’s most watched announcements. Second, Russia’s position in the fuel cycle (Section 2.6) — leverage no miner has. Third, Western policy: the US ban on Russian enriched uranium and the Euratom Supply Agency’s role in co-signing every EU utility’s purchase contract, which it uses to push diversification. The bodies that publish the market’s data — the OECD-NEA and IAEA, and the WNA — set no prices.
Table 16. Who shapes the uranium market
| Body or actor | Role | Leverage | Source |
|---|---|---|---|
| Kazatomprom (Kazakhstan) | Largest producer: 25,839 tU on a 100% basis in 2025 | Output guidance; 75% state-owned | Filed | KAP IAR 2025 | production by operation | “Total tU as U₃O₈” 25,839 | p.36; ownership | “62.99336549% belong to Samruk-Kazyna, 12.00663451% – to the Ministry of Finance” | p.45 |
| Rosatom (Russia) | Mining (Uranium One, ARMZ), conversion, 43% of enrichment | Fuel-cycle supply; the US ban’s target | Filed | WNA enrichment page | “Rosatom” 29,133, 2025 |
| Orano (France) | Mining, conversion, enrichment | 90.33% owned by the French state | Filed | Orano AAR 2025 | Shareholders | “90.33% of share capital held by the French State” | p.14 |
| CNNC and CGN (China) | Domestic and overseas mining; enrichment for China’s fleet | State-directed buying for the fastest-growing fleet | Filed | Red Book 2026 | Table 1.19 | “China” 100% government-owned | p.89 |
| Euratom Supply Agency | Co-signs EU utilities’ supply contracts | Diversification policy | Filed | Red Book 2026 | Policy measures in the EU | “all uranium purchase contracts by EU end users … must be concurred by the ESA” | p.137 |
| US Department of Energy | Waivers to the ban on Russian low-enriched uranium (LEU); surplus HEU | Policy and government stocks | Filed | WNA enrichment page | “waivers which the Department of Energy may grant” |
| OECD-NEA and IAEA | The Red Book: resources, production, demand | Data and projections; no price role | Filed | Red Book 2026 | title page | “Uranium 2026: Resources, Production and Demand” | p.1 |
| World Nuclear Association | Industry data and the Nuclear Fuel Report | Data; no price role | Filed | WNA fuel report page | “biennial report on the outlook for nuclear power and nuclear fuels” |
Source: as cited row by row — Kazatomprom’s 2025 Integrated Annual Report, Orano’s Annual Activity Report 2025, the WNA’s Uranium Enrichment and World Nuclear Fuel Report 2025 pages and OECD-NEA/IAEA, Uranium 2026. The Leverage column is this guide’s judgement.
2.8 Strategic reserves & inventories
Uranium has no strategic petroleum reserve, but it has stockpiles that have shaped the market for thirty years. The largest disclosed stocks are commercial: the EU and US utility inventories of Table 12, which together hold about 88,000 tU — well over a year of world requirements. Government stocks are larger and less visible: the Red Book notes that the US and Russian governments still hold significant natural and highly enriched uranium (HEU) from military programmes, though no official figures exist, and the US National Nuclear Security Administration has run an HEU down-blending programme whose contract was extended through 2027. The great past release was Megatons to Megawatts, whose 500 tonnes of Russian weapons HEU fed US reactors until 2013. Two fuel banks add an emergency backstop: a 120-tonne low-enriched uranium reserve at Angarsk in Russia and the IAEA’s 90-tonne LEU Bank in Kazakhstan, operational since 2019. China, the fastest-growing consumer, publishes no inventory figures, but the Red Book cites a WNA estimate, built from import statistics, of about 147,000 tU at 1 January 2025 — more than the EU and US together — and about 15,000 tU for India. The through-line: the stocks that bridged two decades of mine shortfalls are finite, and the Red Book expects secondary supply to keep declining.
Source: OECD-NEA/IAEA, Uranium 2026, “Natural and enriched uranium stocks and inventories”, “HEU from Russia”, “HEU from the United States” and “Fuel banks”, pp.128–132, and “Supply and demand to 2050”, p.139; the WNA’s Military Warheads as a Source of Nuclear Fuel page. Inventories from Table 12.
3. Nuclear reactors & the fleet
Uranium demand begins and ends with the machines that burn it. Section 2.2 counted the fleet; this section opens the box — what the reactors are, how long they run and what the next generation will need — because reactor design decides how much uranium a plant uses, at what enrichment and for how many decades, the three variables behind the demand curve.
3.1 Reactor types & how they work
A reactor controls the chain reaction of Section 1.1. A moderator slows the neutrons released by splitting U-235 so they split more atoms; a coolant carries the heat away to raise steam. The choice of moderator and coolant defines a reactor’s type. The workhorse is the pressurised water reactor (PWR), 74% of the 413 reactors operating at the end of 2025: ordinary (“light”) water is both moderator and coolant, held at about 150 times atmospheric pressure so it reaches some 325 °C without boiling, and it burns uranium enriched to 3–5%. The boiling water reactor (BWR), 10%, lets the water boil in the core and drives the turbine directly. The pressurised heavy-water reactor (PHWR, or CANDU), 11%, uses heavy water, a moderator efficient enough to run on natural, unenriched uranium — no enrichment at all. The rest are Britain’s gas-cooled reactors and Russia’s graphite-moderated RBMKs, both being retired, plus two fast reactors and one high-temperature gas reactor.
Type matters to the market because it sets the fuel specification. Light-water reactors — 84.5% of the fleet — are what make enrichment the strategic choke point; CANDUs buy natural uranium only. Most advanced designs need fuel enriched beyond 5% (Section 3.4).
Table 17. World reactor fleet by type, end-2025
| Reactor type | Moderator / coolant | Operational reactors | Share of fleet | Fuel | Main countries |
|---|---|---|---|---|---|
| Pressurised water (PWR) | Light water / light water | 306 | 74.1% | Enriched, 3–5% U-235 | United States, France, China, Russia, South Korea |
| Pressurised heavy water (PHWR) | Heavy water / heavy water | 46 | 11.1% | Natural uranium | India, Canada |
| Boiling water (BWR) | Light water / light water | 43 | 10.4% | Enriched, 3–5% U-235 | United States, Sweden |
| Gas-cooled (GCR) | Graphite / CO₂ | 8 | 1.9% | Enriched | United Kingdom |
| Light-water graphite (LWGR, RBMK) | Graphite / light water | 7 | 1.7% | Enriched | Russia |
| Fast breeder (FBR) | None / sodium | 2 | 0.5% | Enriched; MOX | Russia |
| High-temperature gas (HTGR) | Graphite / helium | 1 | 0.2% | Enriched TRISO | China |
| Total | 413 | 100% |
Notes to Table 17
- The count excludes 23 reactors in suspended operation (17 BWRs, 4 PWRs and 2 PHWRs — 19 of them in Japan); the Red Book’s 441 includes them.
Source: IAEA, Nuclear Power Reactors in the World, 2026 edition, Table 2 “Type and net electrical power of operational reactors, 31 Dec. 2025”, row “TOTAL”, p.11, and Table 23 “Reactor types: abbreviations and summary”, p.77. Shares are computed here. Moderator, coolant and fuel descriptions follow the WNA’s Nuclear Power Reactors page (“Water in the reactor core reaches about 325°C”; PHWRs “generally use natural uranium”); the Main countries column is this guide’s summary of Table 2.
Figure 13. World reactor fleet by type, end-2025 (% of operational reactors)
Figure data: Table 17; “Other” is the GCR, LWGR, FBR and HTGR rows together.
3.2 Reactor generations, Gen I to Gen IV
Reactors are grouped into generations. Generation I were the 1950s–60s prototypes, now retired. Generation II — the PWRs, BWRs, CANDUs and RBMKs built from the 1970s to the 1990s — are the bulk of today’s fleet, designed for a nominal 40-year life. Generation III and III+ are the evolutionary designs built since the 1990s and being built now, with standardised designs and, in III+, passive safety that cools the core without power or operator action; most of the PWRs under construction, including China’s Hualong One (HPR1000) and Korea’s APR-1400 in the IAEA’s list, belong here. Generation IV is the research frontier: an international forum is developing six designs, four of them fast-neutron reactors, cooled by sodium, lead, gas or molten salt, all running hotter than today’s plants. Fast reactors can breed new fuel from U-238 and burn long-lived waste, wringing far more energy from each tonne of uranium.
Table 18. Nuclear reactor generations
| Generation | Era | Defining feature | Examples |
|---|---|---|---|
| Gen I | 1950s–1960s | First prototypes | Shippingport, early Magnox (retired) |
| Gen II | 1970s–1990s | Today’s operating fleet; 40-year design life | PWR, BWR, CANDU, RBMK |
| Gen III / III+ | 1990s–today | Evolutionary; passive safety in III+ | AP1000, EPR, APR-1400, VVER-1200, Hualong One |
| Gen IV | 2030s onward | Fast or high-temperature; closed fuel cycle | Sodium-, lead- and gas-cooled fast reactors, molten-salt, very-high-temperature reactors |
Source: WNA, Generation IV Nuclear Reactors — “six generation IV nuclear reactor technologies. Four are fast neutron reactors”; models under construction from the IAEA’s Nuclear Power Reactors in the World, 2026 edition, Table 13, pp.28–30. The Era and Defining feature columns are this guide’s summary.
Figure 14. The four reactor generations
Source: Table 18; a conceptual sequence with no values.
3.3 Plant life & licence renewal
A reactor’s lifetime uranium appetite depends on how long it runs, and here the story has shifted in nuclear’s favour. Plants were licensed for 40 years, a regulatory and financing convention rather than a physical limit. In the United States the Nuclear Regulatory Commission (NRC) renews licences in 20-year steps: it has completed 64 initial renewals, to 60 years, and 13 plants (25 reactors) have received a subsequent licence renewal to 80 years, from Turkey Point in 2019 to Hatch in 2026, with more under review. Life extension is the cheapest “new” nuclear capacity there is, and every extended reactor keeps buying uranium for decades longer than older demand models assumed.
The counter-pressure is age. At the end of 2025 the world’s operating reactors had been connected to the grid for 33 years on average, and 45% of them for 40 years or more. Extensions and restarts support demand now; from the 2030s, a wave of retirements will cut it unless new build replaces them — the fleet turnover behind the outlook in Section 6.
Table 19. Reactor operating life & licence renewal (US framework)
| Milestone | Operating life | Status | Source |
|---|---|---|---|
| Original operating licence | 40 years | Set by the Atomic Energy Act | Filed | NRC licence-renewal backgrounder | “to operate for up to 40 years” |
| Initial licence renewal | 40 to 60 years | 64 plant applications completed | Filed | NRC backgrounder | “renewed for an additional 20 years at a time”; NRC initial renewal page | “Completed Applications”, 64 rows |
| Subsequent licence renewal | 60 to 80 years | 13 plants, 25 reactors renewed; 3 under review | Filed | NRC SLR page | “subsequent license renewal (SLR) to be the period of extended operation from 60 years to 80 years”; “Completed Applications” (Turkey Point … Edwin I. Hatch); “Applications Currently Under Review” |
Source: US NRC, Backgrounder on Reactor License Renewal, Status of Subsequent License Renewal Applications and Status of Initial License Renewal Applications and Industry Initiatives; counts of plants and reactors computed here from the completed-applications tables. Other regulators run comparable periodic safety reviews.
Figure 15. Operating reactors by years since grid connection, end-2025
Source: IAEA, Nuclear Power Reactors in the World, 2026 edition, Table 14 “Operational reactors, 31 Dec. 2025”, column “Grid Connection”, pp.31–49 — each of the 413 reactors’ age at 31 December 2025, binned here into ten-year bands; the mean, 32.9 years, is computed from the same dates. Reactors in suspended operation are excluded.
3.4 Small modular reactors & advanced designs
The most talked-about forward theme is the small modular reactor (SMR) — a design of up to about 300 MWe per module, against 1,000 MWe or more for a conventional plant, with microreactors below about 20 MWe. The pitch is to replace giant, bespoke, over-budget projects with factory-built modules shipped to site and added one at a time, sited beside industry, remote grids or data centres. More than 100 designs are in development.
Reality is earlier-stage than the headlines. At the end of 2025 the IAEA listed one SMR-class power reactor in commercial operation, China’s HTR-PM at Shidao Bay, and three under construction: China’s ACP100 (Linglong One), Russia’s lead-cooled BREST-OD-300 and Argentina’s CAREM25. The Western leaders were still listed as planned: GE Vernova Hitachi’s BWRX-300 at Darlington in Canada (four units, construction start expected in 2026) and Clinch River in Tennessee, TerraPower’s Natrium at Kemmerer, Wyoming, and Kairos Power’s Hermes 2 demonstration. The fuel-cycle catch: more than half of SMR designs in development need HALEU — high-assay low-enriched uranium, enriched to between 5% and 20% U-235 — and only Russia and China can make it at scale; commercial supply comes only from Russia’s Tenex. Scaling SMRs would deepen the very dependency Section 2.6 describes. For the uranium market, SMRs are a real but gradual upside on top of the existing fleet.
Table 20. Small modular and advanced reactors in the IAEA’s lists, end-2025
| Design | Developer | Type | MWe net per unit | Status | Source |
|---|---|---|---|---|---|
| HTR-PM (Shidao Bay 1) | Tsinghua University (design) | High-temperature gas | 150 | Operating; grid 2021, commercial 2023 | Filed | IAEA RDS-2 2026 | Table 14 | “SHIDAO BAY-1” HTGR | p.34 |
| ACP100 (Linglong One) | CNNC | PWR | 100 | Under construction since July 2021 | Filed | IAEA RDS-2 2026 | Table 13 | “LINGLONG-1” | p.28 |
| BREST-OD-300 | Rosatom | Lead-cooled fast | 300 | Under construction since June 2021 | Filed | IAEA RDS-2 2026 | Table 13 | “BREST-OD-300” | p.30 |
| CAREM25 | CNEA (Argentina) | PWR | 25 | Under construction since August 2015 | Filed | IAEA RDS-2 2026 | Table 13 | “CAREM25” | p.28 |
| BWRX-300 | GE Vernova Hitachi | BWR | 300 | Planned: Darlington SMR1–4 (construction expected 2026), Clinch River | Filed | IAEA RDS-2 2026 | Table 12 | “DARLINGTON SMR1”, “CLINCH RIVER SMR-1” | pp.26–27 |
| Natrium | TerraPower | Sodium fast | 342 | Planned: Kemmerer Unit 1 | Filed | IAEA RDS-2 2026 | Table 12 | “KEMMERER UNIT 1” | p.27 |
| KP-FHR (Hermes 2) | Kairos Power | Fluoride-salt high-temperature | 25 | Planned: two demonstration units | Filed | IAEA RDS-2 2026 | Table 12 | “HERMES 2 UNIT 1” | p.27 |
Source: IAEA, Nuclear Power Reactors in the World, 2026 edition, Tables 12 (planned), 13 (under construction) and 14 (operational); the SMR definition, design count and HALEU supply from the WNA’s Small Nuclear Power Reactors and High-Assay Low-Enriched Uranium pages (“Commercial supply of HALEU is only available from Russian company Tenex”). Status changes quickly; the IAEA’s list is the one pinned here.
4. The companies & the value chain
4.1 The largest uranium companies
Uranium production is dominated by state-owned producers, and the way to size them up in an evergreen guide is by durable fundamentals — production, reserves and cost — not market capitalisation, which moves daily. By 2024 output the leaders are Kazatomprom (Kazakhstan, mostly ISL), Cameco (Canada, the Athabasca mines and a 40% stake in Inkai), France’s Orano, Rosatom’s Uranium One, China’s CGN and CNNC, Uzbekistan’s Navoi Mining and BHP, whose Olympic Dam yields uranium as a by-product. Of the top ten, only Cameco, Kazatomprom and BHP trade on major exchanges as companies in their own right, which is why listed developers and the physical trust (Sections 4.2 and 5.1) loom so large in uranium portfolios.
Table 21. Largest uranium producers by output and reserves
| Company | Country | Type | Listing | Production, 2024 | Reserves | Reserve life | Source |
|---|---|---|---|---|---|---|---|
| Kazatomprom | Kazakhstan | ISL, state-controlled | Public (LSE and AIX: KAP) 1 | 12,463 tU | 282.9 thousand tU, attributable (JORC) | 20.9 years | Filed | KAP IAR 2025 | Ore Reserves | “282.9 thousand tU attributable to the Company” | p.41 |
| Cameco | Canada | Conventional, high grade | Public (NYSE: CCJ; TSX: CCO) | 10,193 tU | 433.0 Mlb U₃O₈, its share (NI 43-101) | 16.0 years 2 | Filed | CCJ AR 2025 | Mineral reserves | “Total” 760.8 (100%), 433.0 (our share) | p.110 |
| Orano | France | Diversified fuel cycle | State-owned (unlisted) 3 | 6,815 tU | 101,341 tU, its share | n/d | Filed | Orano AAR 2025 | “Mineral reserves of Orano deposits comprise 101,341 metric tons of uranium” | p.35 |
| Uranium One | Russia | ISL (Kazakhstan), state | State-owned (Rosatom) | 5,829 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
| CGN | China | State | State-owned (CGN) | 5,761 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
| Navoi Mining | Uzbekistan | ISL, state | State-owned | 4,000 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
| CNNC | China | State | State-owned | 3,286 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
| Rosatom (ARMZ) | Russia | State | State-owned | 2,738 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
| BHP | Australia | By-product of copper | Public (ASX, LSE, NYSE: BHP) | 2,693 tU | Olympic Dam: 591 Mt at 0.57 kg/t U₃O₈ (JORC) 4 | n/d | Filed | BHP AR 2025 | Ore Reserves | “Olympic Dam” UG Sulphide 591 Mt, 0.57 kg/t U₃O₈ | p.210 |
| General Atomics / Quasar | United States / Australia | ISL (Beverley, Four Mile) | Privately held | 1,808 tU | n/p | n/d | Filed | WNA mining production page | “Uranium production by company 2024”; reserves not published |
Notes to Table 21
Reserves are as each company filed them at the end of 2025.
- Kazakhstan’s state owns 75% (Samruk-Kazyna 62.99%, the Ministry of Finance 12.01%); 25% trades as shares and global depositary receipts on the Astana International Exchange and the London Stock Exchange (KAP IAR 2025, pp.45–46).
- Cameco’s reserve life excludes Inkai on both sides, because Cameco books its Inkai share as purchases, not production: (433.0 − 96.5 Mlb of Inkai) ÷ 21.0 Mlb produced in 2025 = 16.0 years (CCJ AR 2025, pp.57, 110). Kazatomprom’s is attributable reserves ÷ 13,519 tU of attributable 2025 production (p.35).
- 90.33% owned by the French state (Orano AAR 2025, p.14).
- BHP prints ore tonnage and grade, not contained uranium; a further 43 Mt of low-grade ore grades 0.28 kg/t. Uranium is recovered at about 65%, so a reserve life is not comparable and is not computed.
Source: production — WNA, World Uranium Mining Production, table “Uranium production by company 2024”, one basis for every row; reserves — the companies’ filings as cited: Kazatomprom 2025 Integrated Annual Report, Cameco 2025 Annual Report, Orano Annual Activity Report 2025, BHP Annual Report 2025. “n/p” = not published in the documents checked; “n/d” = not derivable on a matching basis. Reserve lives are computed here. No market-capitalisation figures are shown by design.
Figure 16. Largest uranium producers, 2024 (tU)
Figure data: Table 21, production column; “All others” is the WNA’s “Other” row, 4,627 tU.
4.2 Company archetypes along the value chain
Uranium exposure spans a wide risk spectrum, and an investor should match the archetype to the goal. Explorers offer option value and no cash flow. Developers and restarts are permitting or building mines — in the Athabasca Basin, Denison’s Phoenix ISL project (56.7 Mlb of reserves at 11.7% U₃O₈) received its construction licence in February 2026 and targets first production by mid-2028, while NexGen’s Arrow (239.6 Mlb at 2.37%) awaits its regulator’s decision; value is unlocked at first production, with high leverage to price. Producers sell uranium with direct exposure, shaped by cost-curve position and contract book. Physical holders own pounds of U₃O₈ and track spot. Fuel-cycle companies — converters and enrichers — earn service margins tied to policy as much as to the uranium price. Royalty companies hold a share of others’ output. Among the companies analysed on this blog are the explorer IsoEnergy, the producer Boss Energy, the enricher Centrus Energy and the royalty company Uranium Royalty Corp.
Table 22. Uranium company archetypes
| Archetype | What they do | Revenue model | Examples | Price sensitivity |
|---|---|---|---|---|
| Explorer | Search for deposits | None (raise and spend) | Early-stage Athabasca juniors | Very high |
| Developer / restart | Build or re-open mines | None until production | Denison (Phoenix), NexGen (Arrow) | High |
| Producer | Mine and sell uranium | Uranium sales less cost | Kazatomprom, Cameco, Paladin | High |
| Physical holder | Hold pounds of U₃O₈ | None; tracks spot via NAV | Sprott Physical Uranium Trust | One-to-one with spot |
| Royalty | Finance mines for a share of output | Royalty income | Uranium Royalty Corp. | Medium |
| Fuel cycle (convert, enrich) | Process uranium | Service fees per kgU or SWU | Urenco, Orano, Centrus | Low; policy-driven |
Source: developer reserves and status from Denison’s 2025 Annual Information Form (“Total Proven and Probable”, p.38; “Phoenix remains on track for first production by mid- 2028”, p.8) and NexGen’s 2025 Annual Information Form (“Probable Mineral Reserves were estimated at 240 Mlb U₃O₈ contained in 4.6Mt grading 2.37% U₃O₈”, p.3; the CNSC decision “remains pending as of March 3, 2026”, p.7), each also filed on SEDAR+. The archetypes, revenue models and sensitivity ratings are this guide’s judgement, following the Metal Pilot project-type taxonomy.
4.3 Infrastructure & balance-sheet assets
What a uranium company owns decides what its filings tell you. A producer’s balance sheet rests on reserves and resources (in tU or Mlb U₃O₈ at a % U₃O₈ grade, reported under JORC, NI 43-101 or S-K 1300), its mines — ISL wellfields, or pits and underground workings feeding a mill — and, for integrated companies such as Cameco and Orano, conversion and enrichment plants. A uranium-specific asset is the contract book: the long-term sales contracts and the prices or formulas locked into them, which set near-term revenue more than spot does. Cameco’s realised price in 2025 was $62.11/lb against a spot average of $73.54. And watch gross versus net: Cigar Lake, McArthur River, Inkai and most Kazakh mines are joint ventures, so only the attributable share reaches shareholders — Kazatomprom’s attributable output was 13,519 tU of the 25,839 tU its mines produced in 2025.
Table 23. Uranium-company asset types and metrics
| Asset type | What it does | Key metric | Unit |
|---|---|---|---|
| Reserves and resources | The in-ground uranium base | Tonnage, grade, contained uranium | Mt; % U₃O₈; Mlb or tU |
| Mines (ISL or conventional) | Extract uranium | Annual output; recovery | tU/yr or Mlb/yr; % |
| Mill or processing plant | Produce yellowcake | Throughput, licensed capacity | Mlb U₃O₈/yr |
| Conversion and enrichment | Fuel-cycle services | Capacity | tU/yr; thousand SWU/yr |
| Contract book | Locked-in sales | Committed volume; realised price | Mlb; $/lb |
| Joint-venture interests | Shares of others’ mines | Ownership %; attributable output | %; tU |
Source: this guide’s summary; the examples in the text are from Cameco’s 2025 Annual Report (“Average realized price (US$/lb)" 62.11, p.57) and Kazatomprom’s 2025 Integrated Annual Report (“Production volume of U₃O₈ (attributable basis)" 13,519 tU and (100% basis) 25,839 tU, p.35).
5. Investing in uranium
5.1 Ways to own it — exposure vehicles
Uranium’s menu of vehicles is unlike any other commodity’s, because a private investor cannot store yellowcake: only licensed facilities may hold it. The market’s answer is the physical trust. The Sprott Physical Uranium Trust, listed in Toronto since July 2021, is a closed-end trust that holds substantially all its assets in U₃O₈ stored at the converters of Cameco, ConverDyn and Orano — the uranium equivalent of a bullion fund, and since 2021 a buyer of physical pounds on the spot market. Read the structural catch first: a closed-end trust’s units trade at a premium or discount to the net asset value of the uranium it holds, so a buyer who pays a premium can lose money as it narrows even while uranium rises. The trust publishes the gap daily; check it before buying.
Futures exist — CME’s UxC U₃O₈ contract, 250 lb, cash-settled — but are thin, and no futures-based fund is offered to retail investors, so the contango roll cost that erodes oil and gas ETFs does not arise; the trust’s premium or discount is uranium’s equivalent trap. Mining equities are the leveraged route, from Cameco and Kazatomprom through developers to explorers, and sector equity ETFs bundle them. Fuel-cycle companies — converters and enrichers — give fee-based exposure to the move away from Russian supply rather than to the uranium price.
Table 24. Ways to own uranium
| Vehicle | What you actually own | Tracks spot? | Key structural cost or risk | Best suited for |
|---|---|---|---|---|
| Physical trust | Pounds of U₃O₈ in licensed storage | Yes, through NAV | Premium or discount to NAV | Direct exposure, patient capital |
| Futures (CME UxC U₃O₈) | Cash-settled contracts on UxC spot | Near-term | Thin liquidity; margin calls | Professionals hedging |
| Producer equities | Mining businesses and their contract books | No — amplifies | Contract books delay the benefit of a rally | Leverage to the price |
| Developer and explorer equities | Projects awaiting finance or licences | No — amplifies more | Financing, permitting, execution | High-conviction leverage |
| Sector equity ETFs | A basket of uranium-related shares | No — blended | Concentration in a few large names | Diversified equity exposure |
| Fuel-cycle equities | Conversion and enrichment margins | Weakly | Policy and plant-build risk | The de-Russification theme |
Source: Sprott, Sprott Physical Uranium Trust fund page — “Closed-End Trust”; “invests and holds substantially all of its assets in uranium in the form of U₃O₈”; storage at “Cameco (Canada); ConverDyn (U.S.); Orano (France)"; inception “July 19, 2021”; and CME Group, UxC Uranium U₃O₈ Futures Contract Specs. Structural attributes only — no fees, premiums or prices, which change daily; the ratings are this guide’s judgement.
5.2 How to value & screen uranium miners
The facts above turn into a repeatable checklist. For a producer: the reserve base — how many pounds, at what grade (Athabasca’s 6–16% U₃O₈ and Kazakh ISL’s 0.03% are the two ends of the quality spectrum); the cash cost and all-in sustaining cost (AISC) per pound — who keeps mining through a downturn; and, uniquely for uranium, the contract book — how much output is sold forward and on what terms, since a producer fully committed at old prices captures little of a rally. For developers: the project’s net present value and internal rate of return at a conservative price, the capital cost, and the incentive price that would justify building it. For explorers: grade, drill results and jurisdiction.
The most useful single tool is the cost curve: rank producers from cheapest to dearest, draw the price across it, and see who earns a margin and who needs a higher price to exist. Few uranium producers publish comparable unit costs; the three that do, for 2025, sit far below the spot price — Kazatomprom’s attributable C1 cash cost was $18.06/lb, Cameco’s cash cost of produced pounds about $17, and Paladin’s restarted Langer Heinrich $40.20. To run the same screen — resources, cost and contract cover — across every listed uranium company, use the Metal Pilot uranium dataset.
Table 25. Uranium-miner screening metrics
| Metric | What it tells you | Good vs. concerning | Where to find it |
|---|---|---|---|
| Reserves and grade | Scale and ore quality | Larger and higher-grade is better | Reserve statement (JORC, NI 43-101, S-K 1300) |
| Cash cost / AISC ($/lb) | Cost competitiveness | Well below the long-term price | Annual report, MD&A |
| Contract book | Exposure to a rising or falling price | Some uncommitted volume is upside | MD&A, contract-portfolio notes |
| Realised vs. spot price | How fast price moves reach revenue | A small lag means more torque | MD&A |
| Incentive price | Price needed to build new supply | Below the long-term price = buildable | Feasibility study, company guidance |
| Jurisdiction and permitting | Political and licensing risk | Established regimes are safer | Country profile, regulator |
Source: this guide’s judgement, using the cost and price lines reported in the producers’ filings (Table 26) and the reporting codes in Table 21.
Table 26. Cash cost of three uranium producers per pound of U₃O₈, 2025
| Producer | Operation | Cost measure | Cost | Output | Source |
|---|---|---|---|---|---|
| Cameco | McArthur River / Key Lake, Cigar Lake | Cash cost of produced pounds, its share | $16.98/lb 1 | 21.0 Mlb | Filed | CCJ AR 2025 | Uranium average unit cost of sales | “Cash cost” 23.74 ($/lb) | p.58 Derived | ÷ 1.3978 |
| Kazatomprom | All Kazakh assets | C1 cash cost, attributable | $18.06/lb | 35.15 Mlb | Filed | KAP IAR 2025 | Uranium segment C1 cash cost | “C1 Cash cost (attributable basis) USD/lb” 18.06 | p.37 |
| Paladin | Langer Heinrich (100%), year to June 2025 | Average cost of production | $40.20/lb | 3.0 Mlb | Filed | PDN AR 2025 | Highlights | “US$40.2/lb Average cost of production” | p.10 |
Notes to Table 26
- Cameco reports in Canadian dollars; converted here at the Bank of Canada’s 2025 average of 1.3978 Canadian dollars per US dollar (FXUSDCAD). Its share of output excludes Inkai, which it books as purchases.
- The three measures are not identical: Kazatomprom’s C1 excludes sustaining capital (its AISC was $29.53/lb), Cameco’s excludes non-cash costs (its total production cost was C$35.06/lb), and Paladin’s figure covers a mine in its first full year after restart, against a target of $28.50–31/lb (PDN AR 2025, p.63).
Source: the producers’ filings as cited — Cameco 2025 Annual Report, Kazatomprom 2025 Integrated Annual Report, Paladin 2025 Annual Report (ASX announcements → Annual Report 2025); FX from the Bank of Canada’s daily exchange rates, series FXUSDCAD, 2025 mean computed here. Orano, CGN, CNNC, Navoi and Rosatom publish no unit cost, so the curve covers about 34% of the Red Book’s expected 2025 output.
Figure 17. Uranium cost curve of three producers, 2025 ($/lb U₃O₈)
Figure data: Table 26 — bars left to right Cameco, Kazatomprom and Paladin; each bar’s width is its share of the three’s 59.2 Mlb, its height its cash cost. The dashed line is the 2025 average spot price, $73.54/lb (Table 3).
5.3 Macro regimes, rates & correlations
Uranium’s behaviour is the most idiosyncratic of any commodity in this series, and that is its defining investment feature. Demand is set by a fleet that changes over decades and must be fuelled whatever the economy does, so uranium does not trade like “Dr. Copper” or like an inflation hedge; its price answers to reactor policy, producer discipline, the drawdown of stocks, financial buying and fuel-cycle geopolitics. (For how the major commodities compare across macro regimes — a framework uranium largely sits outside — see the macro regime guide.) Interest rates and the dollar matter only at the margin: uranium is priced in dollars, and high rates raise the cost of carrying inventory, but neither has driven a uranium cycle.
Table 27. Uranium across economic regimes
| Regime | Typical behaviour | Why | Historical example |
|---|---|---|---|
| Nuclear-build optimism | Strong | Demand outlook lifts; utilities contract ahead | 2003–2007: annual spot +760% ($11.55 to $99.33) |
| Accident or policy reversal | Very weak | Reactors shut; a glut builds | 2011–2016: annual spot −55%; February 2011 – November 2016 month-end −74% |
| Producer supply discipline | Strong | Mines idled; producers buy spot to cover contracts | June 2018 – May 2020: spot +50%, S&P 500 +6% |
| Financial buying | Strong | Physical buyers absorb thin spot supply | June to September 2021, the trust’s first months: spot +32%, S&P 500 +5% |
| Recession, risk-off | Largely independent | Reactor fuel demand is inelastic | June to December 2008: spot −11%, S&P 500 −35%; February to May 2020: spot +37%, S&P 500 −11% |
| Fuel-cycle supply shock | Strong | Buyers turn away from Russian fuel | February 2022 – January 2024: spot +106%, S&P 500 +9% |
Source: uranium spot — Cameco, Uranium Price (annual averages, Table 3, and month-end prices); S&P 500 — Robert Shiller, ie_data, sheet “Data”, column “P” (monthly average of daily closes). Moves computed here. The Typical behaviour and Why columns are this guide’s judgement; regime descriptions are historical, not predictive.
Past performance. Uranium is a study in extremes. On month-end prices the spot rose 19-fold from $7.10 at the end of 2000 to $136 in June 2007, fell 87% to $18.00 by November 2016, then rose more than fivefold to $100.25 by January 2024. Monthly price changes over 2000–2025 had an annualised volatility of 23.6%, above copper’s 21.1% and nearly double gold’s 12.7%. Adjusted for inflation, the 2025 average price was still less than half the 2007 average. Past performance is not indicative of future results.
Correlations. Over January 2000 to December 2025, monthly changes in the uranium price were almost uncorrelated with mainstream assets — between 0.00 and +0.10 with the S&P 500, copper, Brent and gold. Even in the financial crisis the link was weak (+0.07, September 2008–March 2009). The strongest relationships are with the companies that mine it — +0.33 with Cameco and +0.38 with the Global X Uranium ETF — and those have strengthened since 2013.
Table 28. Uranium correlations, monthly, 2000–2025
| Asset | Correlation, full window | Window | 2000–2012 | 2013–2025 | Note |
|---|---|---|---|---|---|
| Global X Uranium ETF (URA) | +0.38 | Dec 2010–Dec 2025 | n/a 1 | +0.41 | The equity proxy; amplifies |
| Cameco (CCJ) | +0.33 | Feb 2000–Dec 2025 | +0.25 | +0.41 | The largest listed producer |
| S&P 500 | +0.10 | Jan 2000–Dec 2025 | +0.22 | −0.02 | Largely independent |
| Copper | +0.05 | Jan 2000–Dec 2025 | +0.09 | 0.00 | Not a growth-cycle metal |
| Brent crude | +0.02 | Jan 2000–Dec 2025 | +0.04 | −0.01 | Different drivers |
| Gold | 0.00 | Jan 2000–Dec 2025 | −0.08 | +0.08 | No safe-haven link |
Notes to Table 28
- URA began trading in November 2010, so its 2000–2012 sub-window would hold only 24 monthly changes (+0.20 on Dec 2010–Dec 2012) — too short to set beside the other rows.
Source: computed here from monthly percentage changes — uranium from Cameco’s month-end spot price; gold, Brent and copper from the World Bank’s Pink Sheet, monthly, sheet “Monthly Prices”, columns “Gold”, “Crude oil, Brent” and “Copper”; the S&P 500 from Robert Shiller’s ie_data; Cameco and URA month-end closes from CNBC’s chart data (split-adjusted closes, used for percentage moves only, because no primary price history can be downloaded free). Uranium and the equities are month-end values, the others monthly averages. Correlations are time-varying and can break down around shocks; the Note column is this guide’s judgement.
Figure 18. Uranium’s correlations with major assets, monthly, 2000–2025
Figure data: Table 28, full-window column; Brent crude (+0.02) is in the table only. Shading is on a fixed scale where 0.75 or more is solid.
The equity amplifier. A low monthly correlation hides how much the miners amplify the big moves. In the post-Fukushima bear market the annual average spot price fell 55% from 2011 to 2016; Cameco’s shares fell 61% and the Global X Uranium ETF 83%, its junior-heavy basket hit hardest. In the recovery, annual-average spot rose 184% from 2020 to 2024 while Cameco rose 385% — though URA, which also holds non-producers, rose a smaller 163%. Miners are the geared expression of the uranium price, in both directions; for the juniors, gearing includes the risk of not surviving the downturn.
Source: annual averages of Cameco’s month-end spot price (Table 3) against annual averages of daily closes for Cameco (NYSE:CCJ) and the Global X Uranium ETF (URA), CNBC chart data, 2011–2024 — concluded episodes, computed here.
5.4 Price drivers & cycles
Stripped of noise, the uranium price answers to a short list of forces, and the clearest evidence comes from concluded episodes. On the demand side: the reactor pipeline of new builds, restarts and retirements, nuclear policy, and financial buying of physical pounds. On the supply side: producer discipline, the drawdown of secondary supply, and fuel-cycle geopolitics. The recurring pattern is a long lag between price and supply — a new mine takes a decade — which turns modest imbalances into violent cycles.
Figure 19. Nine episodes that shaped the uranium market, 2003–2024
- 2003The 2003–07 boomAnnual average spot climbs from $11.55 to $99.33/lb; the month-end price peaks at $136 in June 2007.
- 2008BustBy December 2008 spot is $52.50/lb, 61% below the peak, while the long-term price holds at $70.
- 2011FukushimaThe March accident idles Japan's fleet; spot drops from $69.63 in February to $52.88 by June.
- 2013Megatons to Megawatts endsDeliveries of down-blended Russian warhead uranium to the United States are completed.
- 2016The troughSpot hits $18.00/lb in November, 74% below February 2011, amid a supply glut.
- 2018Producers cutCameco idles McArthur River until November 2022; producers buy spot to cover their contracts.
- 2020Pandemic curtailmentsMine suspensions lift spot from $24.80 in February to $33.93/lb in May.
- 2021Sprott trust launchedFrom July 2021 a listed trust buys physical U₃O₈, turning investor money into spot demand.
- 2024Away from Russian fuelAfter the 2022 invasion, spot reaches $100.25/lb in January 2024; the US bans Russian LEU from August.
Source: prices from Cameco’s month-end and annual industry averages (Table 3); events from OECD-NEA/IAEA, Uranium 2026, “Uranium prices”, pp.135–137 (Fukushima, the 2018–2019 production cuts, the 2020 curtailments, the 2022 war and the May 2024 Prohibiting Russian Uranium Imports Act) and p.131 (Megatons to Megawatts, completed at the end of 2013); McArthur River’s idling to November 2022 from p.139; the trust’s launch from Sprott’s fund page; the August 2024 start of the US ban from the WNA’s Uranium Enrichment page.
The settled case studies show each driver at work. The 2003–2007 boom — a nuclear-renaissance narrative, speculation and supply problems at a few critical facilities — carried the annual spot average from $11.55 to $99.33, before buyers balked and the 2008 crisis forced distressed sellers out. The Fukushima accident of March 2011 idled Japan’s fleet and set off a six-year glut that took spot to $18.00 in November 2016. Supply discipline came next: Cameco idled McArthur River from 2018 and producers bought spot to cover their contracts, and the 2020 pandemic shut more mines. The Sprott trust, launched in July 2021, turned investor money into physical demand on a thin spot market. Finally, the 2022 invasion of Ukraine sent Western buyers looking for non-Russian fuel; spot reached $100.25 in January 2024, and the US ban on Russian low-enriched uranium followed in August. The durable lesson: uranium rallies when a growing reactor fleet meets a supply base producers have spent years shrinking, and collapses when an accident or a glut destroys confidence.
Table 29. Uranium price drivers
| Driver | Direction of effect | Why | What to watch |
|---|---|---|---|
| Reactor build and retirement | More reactors, higher | Sets long-run demand | Construction starts, restarts, licence renewals |
| Nuclear policy and sentiment | Pro-nuclear, higher | Drives contracting by utilities | National targets; the COP28 pledge |
| Producer discipline | Cuts, higher | Idled mines tighten a thin market | Kazatomprom and Cameco guidance |
| Secondary supply | Drawdown, higher | The buffer behind the primary gap is finite | Utility inventories (Table 12); underfeeding |
| Financial buying | Inflows, higher | Physical buyers absorb spot supply | The physical trust’s holdings and premium |
| Fuel-cycle geopolitics | Disruption, higher | Russia holds 43% of enrichment | US waivers; Western capacity additions |
Source: this guide’s synthesis of the episodes above and Tables 10, 12 and 15; directions are historical tendencies, not forecasts.
Those drivers resolve into a simple forward frame. It is not a forecast — uranium’s history warns against those — but a set of conditions and the price regime each would produce.
Table 30. Uranium price scenarios (illustrative, not forecasts)
| Scenario | Key conditions | Likely price regime |
|---|---|---|
| Bull | Requirements track the Red Book’s high case; life extensions and restarts continue; secondary supply runs down; new mines and Western enrichment lag | Sustained strength; long-term contract prices lead spot higher |
| Base | Requirements between the Red Book’s low and high cases; restarts and committed mines roughly keep pace until the mid-2030s | Firm long-term prices; spot volatile around them |
| Bear | A serious accident or policy reversal; faster-than-expected new supply; financial holders sell | A sharp, multi-year de-rating, as after 2011 |
Source: this guide’s illustrative frame, conditioned on the drivers in Table 29 and the OECD-NEA/IAEA Uranium 2026 scenarios in Table 31; the rows describe conditions, not forecasts.
5.5 Risks, controversies & ESG
The bull case has real counterweights, and uranium’s are unusually binary. The dominant risk is a nuclear accident: Chernobyl (1986) and Fukushima (2011) each triggered policy reversals that cut demand for years. Policy reversal more broadly — Germany’s completed exit, from 169.6 TWh of nuclear output in 2000 to none in 2024 — can remove demand at a stroke. Resource nationalism bites in a market this concentrated: Niger’s 2023 coup ended in the 2025 nationalisation of Orano’s Somaïr mine and the loss of its Imouraren permit. And the deepest structural risk is dependence on Russian enrichment, which the West is replacing slowly and expensively.
Uranium mining also carries distinctive ESG exposure: the radioactivity of ore, waste rock and mill tailings; groundwater protection at ISL wellfields; the unresolved question of long-term spent-fuel disposal; and weapons proliferation, policed by IAEA safeguards. Against that, nuclear’s defenders point to an energy source that is low-carbon, always on and compact. Substitution is real at the margin — renewables and storage compete for new generation — but for an existing reactor uranium is irreplaceable, which is what keeps demand inelastic. These are contested questions, and reasonable analysts weigh them differently.
Figure 20. Uranium risk map — likelihood vs. impact
Source: this guide’s qualitative assessment of the risks above; positions are judgement, not data. Germany’s generation is from the Energy Institute’s Statistical Review of World Energy 2026 — data workbook, sheet “Nuclear Generation - TWh”, row “Germany”; Niger from the Red Book, Table 1.17, p.77, and “Supply and demand to 2050”, p.139, and Orano’s Annual Activity Report 2025, p.36.
6. Future outlook & forecasts
Uranium demand is unusually forecastable — it is set by a fleet that changes slowly and visibly — and both the agency and the industry project growth. Forecasts remain scenarios: the near-term ones rest on reactors already operating or under construction; the long-term ones on policy.
6.1 Demand
The Red Book projects world reactor requirements rising from 64,584 tU in 2024 to between 84,765 tU (low case) and 143,857 tU (high case) by 2050 — up 31% at the low end and more than doubling at the high end — as nuclear capacity grows from 398 GWe to between 565 and 916 GWe, led by East Asia. The WNA’s World Nuclear Fuel Report 2025 is more aggressive in the near term: from 68,920 tU in 2025 its Reference scenario reaches just over 150,000 tU by 2040, with 746 GWe of capacity, and its Upper scenario more than 204,000 tU. The two bodies count requirements differently, so their levels should not be mixed; what they agree on is the direction.
6.2 Supply and the gap
The catch is supply. On the Red Book’s figures, existing and committed mines could produce 74,209 tU a year in 2030, enough for the low case but not the high; their capability then falls to 37,755 tU by 2040 and 15,525 tU by 2050 as orebodies deplete. Adding every planned and prospective project lifts 2030 capability to 98,589 tU, but even that falls to 57,455 tU by 2050 — below both demand cases from 2040. Closing the gap needs new mines, with long lead times, and a rebuilt Western conversion and enrichment industry, while the secondary stocks that bridged past gaps keep shrinking. Capability is not output — the Red Book notes that actual production “seldom, if ever” matches it (p.92).
Table 31. Uranium requirements and mine production capability to 2050
| Red Book 2026 case | 2024 | 2030 | 2035 | 2040 | 2045 | 2050 |
|---|---|---|---|---|---|---|
| Requirements, low case | 64,584 tU/yr | 66,835 tU/yr | 78,034 tU/yr | 83,001 tU/yr | 84,857 tU/yr | 84,765 tU/yr |
| Requirements, high case | 64,584 tU/yr | 81,768 tU/yr | 92,474 tU/yr | 110,738 tU/yr | 127,884 tU/yr | 143,857 tU/yr |
| Capability, existing and committed mines (A-II) | n/p | 74,209 tU/yr | 57,335 tU/yr | 37,755 tU/yr | 32,695 tU/yr | 15,525 tU/yr |
| Capability incl. planned and prospective (B-II) | n/p | 98,589 tU/yr | 94,075 tU/yr | 76,485 tU/yr | 69,510 tU/yr | 57,455 tU/yr |
| Gap, high case vs. A-II | — | 7,559 tU/yr | 35,139 tU/yr | 72,983 tU/yr | 95,189 tU/yr | 128,332 tU/yr |
| Gap, high case vs. B-II | — | −16,821 tU/yr | −1,601 tU/yr | 34,253 tU/yr | 58,374 tU/yr | 86,402 tU/yr |
Source: OECD-NEA/IAEA, Uranium 2026, Table 2.4 “Annual reactor-related uranium requirements to 2050”, row “World Total”, low and high columns, p.125, and Table 1.23 “World uranium production capability to 2050”, row “Total”, columns A-II and B-II, p.93 (capability from identified resources below $130/kgU); capacity from Table 2.3, p.124. The 2024 requirement is the actual figure. Gaps = requirements − capability, computed here; a negative gap means capability exceeds requirements. Scenario projections, not measured data. The WNA figures in the text are from its World Nuclear Fuel Report 2025 summary page (the full report is sold).
Figure 21. Uranium requirements vs. mine production capability to 2050 (tU/yr)
Figure data: Table 31, the four projection rows; the requirement lines start from the actual 2024 figure.
6.3 Catalysts to watch
The forward watch-list is concrete. Near term: Kazatomprom’s and Cameco’s production guidance (the swing supply), restarts and life extensions in Japan and the United States, flows into the physical trust, and the build-out of Western enrichment before US waivers for Russian supply end at the start of 2028. Over three to ten years: whether the fleet grows along the Red Book’s high case — China’s 36 reactors under construction, the COP28 pledge to triple nuclear capacity by 2050, SMRs and data-centre demand — and whether new mines such as Denison’s Phoenix and NexGen’s Arrow arrive in time. What would confirm the bull thesis: continued build-out and a still-thin supply pipeline. What would break it: a serious accident, a major policy reversal, or new supply arriving faster than expected.
Table 32. Uranium catalyst calendar
| Catalyst | Timing | Why it matters | Watch |
|---|---|---|---|
| Kazatomprom and Cameco guidance | Quarterly | Swing supply | Production and sales guidance |
| Euratom Supply Agency annual report | Annual, mid-year | EU purchases, prices, inventories | Russian share; inventory cover |
| US EIA Uranium Marketing Annual Report | Annual, mid-year | US purchases, prices, inventories | Origins; contracted volumes |
| US waiver authority for Russian LEU ends | Start of 2028 | Forces substitution of Russian SWU | Urenco and Orano additions |
| Denison Phoenix first production | Targeted mid-2028 | First new Athabasca ISL mine | Construction progress |
| NexGen Arrow licence decision | Pending | The largest undeveloped high-grade deposit | Regulator’s decision; financing |
| OECD-NEA/IAEA Red Book | Biennial; next 2028 | Resources, production, demand | Coverage ratio; capability |
Source: the timings are those stated in the documents behind Tables 12, 14, 15 and 22 and in OECD-NEA/IAEA, Uranium 2026 (the Red Book’s 2026 edition followed the 2024 edition); this guide’s selection.
7. Summary
Uranium is the fuel of nuclear power and almost nothing else: its value rests on U-235, its demand on a fleet of 441 reactors that must be fuelled whatever the price. It is priced privately, under long-term contracts and a thin spot market, and best read as an average — spot ran from $8.18 in 2000 to $99.33 in 2007, down to $21.66 in 2017 and back to $85.14 in 2024, $73.54 in 2025 and $85.34 over the 12 months to September 2026. It is mined in a few places — Kazakhstan 38%, Canada 23%, Namibia 12% — mostly by state companies, increasingly by in-situ leach, at grades from 0.03% to 16% U₃O₈. Demand comes from the United States (a third), China (the growth) and France, and has run ahead of mine supply every year since the 1990s; inventories, recycled material and ex-weapons uranium have filled a gap that was 26% of needs in 2020 and 4–10% in 2024. The choke point is downstream: Russia holds 43% of enrichment capacity, and the West is only now rebuilding its own. The companies — Kazatomprom, Cameco, Orano and a group of state miners — are best compared on production, reserves and cost; the three that publish comparable costs produce far below the spot price. Uranium’s regime is its own: almost uncorrelated with stocks, oil, copper or gold, it rewards reactor build-out, producer discipline and fuel-cycle stress and collapses after accidents, and the miners amplify both. The single variable to watch is the gap between a growing fleet and a mine pipeline that, on the Red Book’s figures, shrinks after 2030.
To go from this big-picture view to the actual companies — every uranium producer and developer screened by resources, cost and contract cover — explore Metal Pilot’s uranium mining stock screener.
Terms in this post (44)
- AISC (all-in sustaining cost)
- The cost to produce an ounce or pound and keep the mine running: operating costs, royalties, sustaining capital and corporate overhead, net of by-product credits, on the World Gold Council’s non-GAAP definition. Lower is better. In depth: Gold market guide — How to value & screen gold miners
- Attributable
- A company’s ownership share of a jointly or partly owned asset’s output or reserves, after every layer of the structure — the part that belongs to its shareholders.
- Brent
- The North Sea crude benchmark that prices most of the world’s internationally traded oil. In depth: Oil market guide — Pricing & benchmarks
- By-product
- A metal recovered incidentally while mining another. Its supply follows the host metal’s output, not its own price.
- C1 / cash cost
- The direct cost of producing a pound or ounce — mining, processing, site overhead and freight, net of by-product credits — before depreciation, royalties and sustaining capital. Lower is better. In depth: Copper market guide — How to value & screen copper miners
- C1 cash cost / AISC
- Cash cost per pound / cash cost plus sustaining capital. Who survives a downturn.
- CME UxC U₃O₈ futures
- Cash-settled uranium futures on CME, settled against UxC’s month-end spot price for U₃O₈ — uranium oxide concentrate, or yellowcake — in dollars a pound. In depth: Uranium market guide — Pricing & benchmarks
- Contango roll cost
- The loss a futures fund takes when it sells an expiring contract to buy a dearer later one. Absent in uranium: no futures fund.
- Contract book
- A producer’s long-term sales contracts. Delays or caps the benefit of a rally.
- Conventional mining
- Open-pit or underground mining with a mill. One of two main supply routes.
- Conversion (UF₆)
- Turning yellowcake into uranium hexafluoride gas. A separate, concentrated fuel-cycle step.
- Enrichment
- Raising U-235 to 3–5% (or more). The choke point: Russia holds 43%.
- Fuel fabrication
- Making pellets and fuel assemblies. The last step before the reactor.
- Grade
- The concentration of metal in the ore — grams a tonne for gold and silver, per cent for base metals and uranium. A higher grade means more metal per tonne mined and usually a lower unit cost. In depth: Gold market guide — What gold is — physical basics & quality
- HALEU
- High-assay low-enriched uranium, 5–20% U-235. Needed by most advanced reactors; Russian-supplied.
- HEU downblending
- Diluting weapons-grade uranium into reactor fuel. The Megatons to Megawatts legacy.
- Identified resources (RAR + inferred)
- Delineated uranium recoverable below a cost threshold. Uranium’s endowment measure.
- In-situ leach (ISL / ISR)
- Dissolving uranium underground and pumping it to surface. 55% of supply; low cost; needs constant drilling.
- Incentive price
- The price needed to justify a new mine. Where new supply comes from.
- Licence renewal (initial / subsequent)
- A regulator’s extension of an operating licence: 40 — 60, then 60 — 80 years in the US. Keeps old reactors buying fuel.
- Low-enriched uranium (LEU)
- Uranium enriched to less than 20% U-235; reactor fuel is typically 3–5%. The fuel of today’s reactors.
- Mlb (million pounds)
- Million pounds — the unit uranium, as U₃O₈, and North American copper output and reserves are reported in.
- Moderator / coolant
- Material that slows neutrons / fluid that removes heat. Defines a reactor type and its fuel.
- MOX / reprocessed uranium
- Fuel recycled from spent fuel. Small; under 1% of requirements.
- Net position
- Mine production minus reactor requirements. Who exports and who imports.
- A listed vehicle holding U₃O₈ / the value of its holdings per unit. Trades at a premium or discount to NAV.
- Price-reporting agency
- A firm, such as Platts or Argus, that assesses and publishes physical commodity prices from reported trades.
- Production
- The metal, oil or gas the company produced in the period at its own share of each mine or field — ounces, pounds or tonnes, or barrels of oil equivalent a day — before or after royalties as it reports; for a royalty or stream holder, its share of the operators’ output. In depth: The Oil & Gas Value Chain: Upstream — Units & reading a production number
- Production capability (A-II / B-II)
- The Red Book’s estimate of what existing and committed (A-II), or also planned and prospective (B-II), mines can produce. The supply side of the outlook.
- PWR / BWR / PHWR
- Pressurised water, boiling water and heavy-water reactors. Light-water types need enriched fuel; PHWRs do not.
- Reactor generations (II, III, IV)
- The fleet built since the 1970s, today’s evolved designs, and the research frontier. Sets fuel type and build risk.
- Reactor requirements
- Uranium a fleet needs in a year. The core demand number.
- Reserve life
- Years the reserves last at the current or guided production rate: reserves divided by a year’s output, on one reserve category. In depth: The Oil & Gas Value Chain: Upstream — How to read & screen a producer
- Reserves
- What a company can show it will produce at a profit from known deposits or fields, at stated prices and costs, estimated under a disclosure code — NI 43-101, JORC or S-K 1300 for mines; SEC rules or NI 51-101 for oil and gas. In depth: The Oil & Gas Value Chain: Upstream — Reserves vs resources
- Royalty
- A share of a mine’s or well’s revenue or profit paid to a government or a royalty holder, ahead of the operator’s own margin. In depth: The Oil & Gas Value Chain: Upstream — Fiscal regimes
- Secondary supply
- Supply from sources other than mines — recycling, inventories and stockpiles.
- SMR
- A small modular reactor — a nuclear reactor of up to about 300 MWe designed to be built in factories.
- Spot vs. long-term price
- One-off delivery vs. multi-year contract reference. Most volume is priced long-term.
- SWU
- Separative work unit: the measure of enrichment effort. Enrichment is priced per SWU.
- tU / lb U₃O₈
- Tonne of uranium metal / pound of uranium oxide; 1 tU = 2,599.8 lb U₃O₈. Agencies count in tU, the market prices in lb.
- tU, kgU (uranium metal)
- Tonnes or kilograms of uranium as metal, the unit fuel services and Kazakh producers report in; one tonne of uranium equals 2,599.79 pounds of U₃O₈.
- U-235 / U-238
- The fissile (0.7%) and the non-fissile (99.3%) isotopes of natural uranium. Why most reactors need enriched fuel.
- Underfeeding / tails
- Using extra enrichment to need less uranium / depleted uranium left after enrichment. Turns spare enrichment into uranium supply.
- U₃O₈ (yellowcake)
- Uranium oxide concentrate, the form mined uranium is sold and priced in, quoted in dollars a pound.
8. Sources, methodology & disclaimer
8.1 Sources, methodology & data vintage
Methodology. Data as of 10 October 2026; the guide is refreshed annually, after the Euratom Supply Agency and US EIA annual reports (mid-year) and each biennial Red Book (the next is expected in 2028). Market structure is for 2024, the latest year both the Red Book and the WNA have counted, with the Red Book’s expected 2025 output alongside. Prices are calendar-year, multi-year and trailing-12-month averages of Cameco’s month-end industry averages of the UxC and TradeTech spot and long-term prices, never single quotes; an annual value is the mean of twelve month-end prices, and real prices deflate by US CPI-U to 2025 dollars. Two mine-production counts are kept apart: the Red Book (61,924 tU in 2024) for countries, ownership, mining methods, resources, requirements by country and the outlook, and the WNA (60,213 tU) for the 2015–2024 world series, the coverage ratio, mines and companies. Correlations use monthly percentage changes, January 2000–December 2025 (the uranium ETF from December 2010); uranium and the equities are month-end values, the other assets monthly averages. Equity episodes use annual averages of daily closes. Every derived figure is computed in scripts kept with this guide’s evidence. Pounds of U₃O₈ convert at 2,599.8 lb per tU and Canadian dollars at the Bank of Canada’s 2025 average. Judgement columns are labelled. Resources, reserves and forecasts are estimates, not measured facts.
Data gaps & substitutions.
- The price. No agency publishes a free monthly series of the trade-press spot price; this guide uses the industry averages Cameco publishes, which match the averages in Cameco’s audited annual report. The IMF’s “Global price of Uranium” (FRED series PURANUSDM) tracks the same spot price until about 2020 but runs well below it since — $69.7 against $85.1 for the 2024 average — so it is not used.
- Requirements by country over time. No agency publishes an annual series of uranium requirements by country back to 2000, so Table 7 shows nuclear generation by region (Energy Institute) as the demand series, and Table 10 implies world requirements from the WNA’s coverage ratio. The Red Book and the WNA count requirements differently (64,584 tU for 2024 against the WNA’s 68,920 tU for 2025).
- Trade. Several producers report no uranium trade to UN Comtrade, so Section 2.6 shows the origins of what EU and US utilities bought (ESA, EIA) in place of exporter and importer tables, and the chokepoint table covers conversion and enrichment, not shipping lanes. The EIA withholds the Russian-origin share of US purchases. The WNA’s latest conversion data are for 2022.
- Companies and mines. Production is the WNA’s 2024 company ranking, one basis for every row. State producers — Uranium One, CGN, Navoi, CNNC, Rosatom — publish no reserves in the documents checked; Orano publishes no group production total; BHP prints no contained uranium. Husab’s and Rössing’s 2025 output are not in the documents checked. Only three producers publish a comparable unit cost, so the cost curve covers about a third of world output, on three definitions.
- Inventories. Only EU and US inventories are published; China’s, Russia’s and India’s are not, and the figures for China and India in Section 2.8 are the WNA estimates the Red Book cites.
- Resources. The WNA’s resource table prints a world total that does not match the Red Book it cites; the Red Book is used directly.
- Exchange and web pages. CME and NRC pages refused scripted downloads and were read in a browser; the WNA, Cameco and Sprott pages are live pages, saved on the date this guide was refreshed.
- Equity prices. No primary daily price history for Cameco or the Global X Uranium ETF can be downloaded free, so the correlations and the equity amplifier (Section 5.3) use CNBC’s chart data — split-adjusted closes, used for percentage moves only.
Refresh log.
- 10 Oct 2026 — full refresh. Moved market structure and the outlook to the Red Book 2026, reactors to the IAEA’s 2026 reactor list, purchases and inventories to the 2025 ESA and EIA reports, and companies and costs to the producers’ 2025 filings. Rebuilt prices on Cameco’s month-end averages and the correlations and equity amplifier on computed monthly data (Cameco’s correlation with uranium is +0.33, not ≈ +0.8; its 2011–2016 fall −61%, not −75%). Corrected the enrichment, reactor-type, ISL and resource figures, removed claims no source supports, and pinned every figure to its document.
- 5 Jul 2026 — first published; revised 18 Jul 2026 (data as of 2 July 2026).
The register below is the guide’s reference list: one band per document, in order of first use, each linked to the edition cited.
Table 33. Source register — the document behind every table and figure
| Table / figure | What it takes | Locator | Period |
|---|---|---|---|
| OECD-NEA/IAEA, Uranium 2026: Resources, Production and Demand (“Red Book”) — printed page numbers | |||
| Figure 1; Tables 4, 8, 11; Figures 4, 9 | Mine production by country and world | Table 1.18 “Historical uranium production”, pp.77–78 | 2022–2025e |
| Table 4; Section 2.1 | Ownership; changes by country; Niger | Tables 1.17, 1.19, pp.77, 89 | 2022–2024 |
| Figure 1; Tables 6, 11; Figures 5, 9 | Reactors and 2024 requirements by country | Table 2.1 “Nuclear power plant data summary”, p.106 | 2024 |
| Figure 1; Table 10 | Production against requirements, 96% | “Primary sources of uranium supply”, p.127 | 2024 |
| Figure 1; Table 9 | Identified recoverable resources by country | Table 1.2a, column "<USD 130/kgU”, pp.24–25 | 1 Jan 2025 |
| Tables 1, 33 | Unit conversions; glossary definitions | Appendix 3, p.539 | — |
| Table 7 note | Nuclear generation, world | Table 2.2, p.107 | 2024 |
| Tables 12, 13; Sections 1.3, 2.5, 2.8 | EU and US utility inventories; China and India inventory estimates; secondary supply, HEU, MOX, RepU, fuel banks; spot share of transactions | Table 2.5, p.131; pp.128–133; p.136 | 2015–2024 |
| Table 13; Figure 10 | Production by mining method | Table 1.22, p.92 | 2022–2025p |
| Tables 16, 31; Figures 19, 21; Sections 2.4, 5.4, 6 | Market developments; ESA role; requirements and capability to 2050; capacity | pp.135–139; Tables 2.3, 2.4, p.124–125; Table 1.23, p.93 | 2024–2050 |
| Cameco, Uranium Price — month-end industry averages of UxC and TradeTech prices | |||
| Figure 1; Tables 2, 3, 27, 28; Figures 3, 18, 19 | Spot and long-term prices; trailing 12 months; episodes; correlations | “Uranium Spot Price”, “Long-term Uranium Price”, month-end | 2000–Sep 2026 |
| WNA, Uranium Enrichment (capacity from the WNA Nuclear Fuel Report 2025) | |||
| Figure 1; Tables 15, 16; Figure 12; Sections 1.1, 2.6 | Enrichment capacity by operator; US ban and waivers; capacity additions; isotopes | Table “World enrichment capacity”; text | 2022, 2025, 2030 |
| WNA, Conversion and Deconversion (from the WNA Nuclear Fuel Report 2023) | |||
| Figure 2; Table 15 | Conversion capacity and production by company | Table “Estimated world primary conversion capacity and UF₆ production 2022” | 2022 |
| US EIA, 2025 Uranium Marketing Annual Report | |||
| Tables 1, 2, 12, 14, 15; Figure 11 | Purchases by origin and price; SWU purchases and price; inventories by owner | Tables 3, 16, 23, pp.21, 45, 59 | 2021–2025 |
| FRED, CPIAUCSL (US CPI-U) | |||
| Table 3 | Deflator for the real price row | Monthly series, annual means | 2000–2025 |
| CME Group, UxC Uranium U₃O₈ Futures Contract Specs | |||
| Tables 2, 24 | Contract unit and settlement | “Contract Unit”, “Settlement Method” | — |
| Euratom Supply Agency, Annual Report 2025 | |||
| Tables 2, 12, 14, 15; Figure 11 | EU prices; purchases by origin; enrichment origin; inventories and requirements | p.11; Table 3, p.12; Table 5, p.15; pp.10, 16–17 | 2025 |
| Cameco, 2025 Annual Report | |||
| Table 3 note; Tables 5, 21, 23, 26; Figures 16, 17 | Price averages; mine output and grades; reserves; realised price; cash cost | pp.57–58, 78, 82, 110 | 2024–2025 |
| Kazatomprom, 2025 Integrated Annual Report (annual reports → 2025) | |||
| Tables 5, 16, 21, 23, 26; Figure 17 | Output by asset; attributable output; reserves; shareholders; C1 and AISC | pp.35–37, 41, 45–46 | 2024–2025 |
| BHP, Annual Report 2025 | |||
| Tables 5, 21 | Olympic Dam uranium output; ore reserves | pp.201, 210 | Year to Jun 2025 |
| Orano, Annual Activity Report 2025 | |||
| Tables 5, 16, 21; Section 5.5 | KATCO operator; state ownership; reserves; Niger | pp.14, 34–36 | 2025 |
| IAEA, Nuclear Power Reactors in the World, 2026 edition (RDS-2) | |||
| Tables 6, 17, 18, 20; Figures 13, 15; Section 3 | Reactors under construction; fleet by type; planned, under construction and operating SMRs; grid-connection dates | Tables 1, 2, 12–14, 23, pp.8–11, 26–49, 77 | 31 Dec 2025 |
| Energy Institute, Statistical Review of World Energy 2026 — data workbook | |||
| Table 7; Figure 6; Section 5.5 | Nuclear generation by region and Germany | Sheet “Nuclear Generation - TWh” | 2000–2025 |
| WNA, World Uranium Mining Production | |||
| Tables 8, 10, 21; Figures 7, 8, 16 | World production and coverage ratio; production by company | Tables “Production from mines (tonnes U)”, “Uranium production by company 2024” | 2015–2024 |
| WNA, In Situ Leach Mining of Uranium | |||
| Table 13; Section 1.1 | ISL share in 2000 | Text, “has risen significantly from 16% in 2000” | 2000 |
| WNA, Military Warheads as a Source of Nuclear Fuel | |||
| Section 2.8 | Megatons to Megawatts | “Megatons to Megawatts, the Russian HEU deal” | 1993–2013 |
| WNA, World Nuclear Fuel Report 2025 (summary page) | |||
| Table 16; Section 6.1 | Reference and Upper scenarios; 2025 requirements | Summary text | 2025–2040 |
| WNA, Nuclear Power Reactors | |||
| Table 17; Section 3.1 | Moderators, coolants and fuels by type | Text | — |
| WNA, Generation IV Nuclear Reactors | |||
| Table 18; Figure 14 | Six Generation IV designs | Text | — |
| US NRC, Backgrounder on Reactor License Renewal; US NRC, Status of Subsequent License Renewal Applications; Status of Initial License Renewal Applications | |||
| Table 19; Section 3.3 | Original 40-year licence and 20-year renewals; completed renewals | Backgrounder text; “Completed Applications” tables | 2000–2026 |
| WNA, Small Nuclear Power Reactors; High-Assay Low-Enriched Uranium | |||
| Table 20; Section 3.4 | SMR definition and design count; HALEU supply | Text | — |
| Denison Mines and NexGen Energy, 2025 Annual Information Forms | |||
| Tables 22, 32; Section 4.2 | Phoenix and Arrow reserves; licensing status | Denison pp.8, 38; NexGen pp.3, 7 | 2025–2026 |
| Sprott, Sprott Physical Uranium Trust | |||
| Table 24; Figure 19; Section 5.1 | Structure, holdings, storage, inception | Fund page | — |
| Paladin Energy, Annual Report 2025 (ASX announcements → Annual Report 2025) | |||
| Table 26; Figure 17 | Langer Heinrich output and cost | pp.4, 10, 63 | Year to Jun 2025 |
| Bank of Canada, daily exchange rates | |||
| Table 26 | US dollar to Canadian dollar | Series FXUSDCAD, 2025 mean | 2025 |
| Robert Shiller, ie_data | |||
| Tables 27, 28; Figure 18 | S&P 500 monthly | Sheet “Data”, column “P” | 2000–2025 |
| World Bank, Commodity Price Data (Pink Sheet), monthly | |||
| Table 28; Figure 18 | Gold, Brent and copper monthly | Sheet “Monthly Prices” | 2000–2025 |
| CNBC chart data — Cameco (NYSE:CCJ) and Global X Uranium ETF (URA) daily closes | |||
| Table 28; Figure 18; Section 5.3 | Monthly and annual equity prices | Split-adjusted closes | 2000–2025 |
| FRED, PURANUSDM — IMF “Global price of Uranium” | |||
| Section 8.1 (data gaps) | 2024 average, as the series not used | Monthly series, annual mean | 2024 |
| This guide — derived, estimated or illustrative | |||
| Figure 1 (notes 2, 7, 8); Figure 9; Tables 3, 4, 6, 9–12, 14, 15, 21, 26, 31 | Market size, real prices, shares, implied requirements and shortfalls, net positions, cover, reserve lives, gaps | Operations stated in each source line | 2000–2050 |
| Tables 27, 28; Figures 15, 18; Section 5.3 | Episode moves, correlations, volatility, amplifier returns, reactor age bands | Computed from the series above | 2000–2025 |
| Figures 2, 14, 20; Tables 16, 18, 22–25, 29, 30, 32; the Terms list | Fuel cycle, generations, risk map, bodies, archetypes, assets, vehicles, screening, drivers, scenarios, catalysts, the Terms list’s wording | This guide’s summaries and judgements | — |
| Tables 2, 11, 13, 16, 17, 22, 24, 27, 28 | The role, leverage, price-response, sensitivity, regime and note columns | Judgement columns, labelled in each source line | — |
Source: compiled from the source lines under each table and figure in this guide; each band links the edition cited, and the closing band lists the figures that are this guide’s own computation or judgement.
8.2 Disclaimer & disclosure
This report is for informational purposes only and is not investment advice, a recommendation, or an offer to buy or sell any security or commodity. Uranium prices are volatile, and the figures here are estimates as of the stated date that will change; resources, reserves, forecasts, correlations and regime descriptions are estimates and historical observations that may not persist. Do your own research and consult a licensed financial adviser before acting. This report was prepared with the assistance of AI; its figures were sourced from the documents in the register above and reviewed, but readers should verify any number before relying on it. The author holds no position disclosed as a conflict in respect of the companies named.