Which countries hold the largest rare earth elements reserves in 2026?

Rare earth elements (REE) reserves by country are dominated by China, which held roughly 44 million metric tons of rare earth oxides as of 2025 according to the United States Geological Survey (USGS). China accounts for close to half of the world's estimated 90-100 million metric tons of identified reserves, and it also processes the overwhelming majority of global rare earth supply. Vietnam ranks second with approximately 22 million metric tons, followed by Brazil with around 21 million metric tons, Russia with roughly 10 million, and India with about 6.9 million metric tons. Australia holds approximately 5.7 million metric tons, representing about 5 percent of global reserves, while the United States holds around 1.9 million metric tons, concentrated almost entirely at the Mountain Pass mine in California.

Reserves, however, are not the same as production or processing capacity. China mined roughly 240,000 metric tons of rare earths in 2024, about 60-70 percent of global mine production, and it refines close to 90 percent of the world's separated rare earths. The United States, despite holding only about 2 percent of reserves, is the second-largest producer thanks to Mountain Pass, but most of its concentrate is still shipped to China for separation. This gap between where reserves sit and where value is added is the central tension in the global rare earth market heading into 2026.

Also worth reading: What is spatial cross-validation in mineral prospectivity mapping and why does it matter for AI-driven rare earth exploration? · How does hyperspectral satellite rare earth detection actually work, and can it find REE deposits from orbit? · How much will rare earth separation technology cost in 2026, and why is it so expensive?

The Global Reserve Rankings in Detail

The USGS Mineral Commodity Summaries, updated each January, remains the most widely cited source for country-level rare earth reserve estimates. Based on the most recent figures and 2025-2026 reporting, the ranking looks like this: China leads with approximately 44 million metric tons, Vietnam holds about 22 million, Brazil about 21 million, Russia roughly 10 million, India approximately 6.9 million, Australia around 5.7 million, the United States about 1.9 million, Greenland roughly 1.5 million, Canada approximately 830,000, and several African nations including South Africa, Madagascar, and Malawi holding smaller but strategically important deposits. Global identified reserves total roughly 90-100 million metric tons of rare earth oxide equivalent.

These numbers deserve skepticism in both directions. Reserve figures are estimates of economically extractable material under current technology and prices, not fixed geological endowments. A country can dramatically increase its reported reserves simply by investing in exploration, as Brazil has done over the past decade, moving from a minor player to the third-largest reserve holder. Conversely, reserves can be downgraded if permitting stalls or economics deteriorate. Vietnam's 22 million ton figure, for example, largely reflects the Dong Pao deposit in Lai Chau province, a single asset whose development has been repeatedly delayed by licensing disputes and policy shifts. Treat every national figure as a snapshot of reported, classified resources rather than an immutable fact.

China: Dominance in Reserves, Production, and Processing

China's position rests on three pillars. First, its 44 million metric tons of reserves include the Bayan Obo deposit in Inner Mongolia, the world's largest rare earth mine, which has been producing since the 1950s and contains a bastnaesite-monazite ore body rich in light rare earths. Second, China produced roughly 240,000 metric tons in 2024, more than the rest of the world combined, under a strict quota system administered by the Ministry of Industry and Information Technology. Third, and most importantly, China controls an estimated 85-90 percent of global separation and refining capacity, plus the majority of rare earth magnet manufacturing, particularly neodymium-iron-boron (NdFeB) magnets used in electric vehicles, wind turbines, and consumer electronics.

This dominance is the product of deliberate industrial policy dating back decades: low prices through the 2000s forced Western competitors like Mountain Pass to shut down, while China consolidated its industry from over 100 small producers into six state-backed groups. Beijing has also tightened export controls. In 2023 it restricted exports of gallium and germanium, and in 2025 it imposed licensing requirements on several heavy rare earth technologies and certain magnet exports, prompting emergency responses from the US, EU, and Japan. Anyone analyzing rare earth reserves by country must therefore separate the geological question (where the deposits are) from the strategic question (who can actually refine and manufacture with them). On the second question, no country comes close to China as of mid-2026.

Brazil: The Fast-Rising Challenger

Brazil has quietly become the story of the decade in rare earth reserves. With approximately 21 million metric tons, it now holds the third-largest reserves globally and the largest outside Asia, according to USGS data and reporting from outlets like DW and ANBA. The country's deposits are geologically favorable: ionic clay deposits in Minas Gerais, similar in type to southern China's heavy rare earth clays, plus monazite-bearing sands along the coast and carbonatite-associated deposits such as Serra Verde, which began commercial production of heavy rare earths in 2024. Serra Verde is currently one of the only significant producers of heavy rare earths outside China, a fact with real strategic weight given that dysprosium and terbium are the elements most exposed to Chinese export restrictions.

Brazil's challenges are equally real. The country lacks domestic separation and refining capacity at scale, meaning most of its output would still need processing abroad unless new facilities are built. Environmental permitting in the Amazon-adjacent states is slow, infrastructure gaps raise logistics costs, and the domestic industry remains fragmented. Brazilian officials and analysts cited in DW's coverage argue the country could cut meaningfully into Chinese dominance within the 2030s, but that timeline depends on billions of dollars in downstream investment that has not yet been committed. Brazil is a reserve giant and an emerging producer, not yet a supply chain alternative.

Vietnam, Russia, India, and Other Major Holders

Vietnam's 22 million metric tons make it the nominal number two, but production remains modest at a few thousand tons annually, constrained by the stalled Dong Pao project and a 2023-2024 government crackdown on illegal mining that disrupted the sector. Vietnam has positioned itself as a partner for Western supply chains, signing cooperation agreements with the United States, but its actual contribution to global supply remains small relative to its reserve base.

Russia's roughly 10 million metric tons include the Tomtor deposit in Yakutia, one of the richest niobium and rare earth deposits known, but Western sanctions since 2022 have effectively removed Russian supply from OECD procurement considerations. India's 6.9 million metric tons are largely monazite beach sands along the southern coast, historically processed under state monopoly through Indian Rare Earths Ltd; recent policy shifts have opened the sector to private players, and Indian media coverage in 2021-2022 highlighted domestic motor and magnet initiatives aimed at reducing import dependence on China.

Australia deserves special mention despite ranking sixth in reserves at about 5.7 million metric tons. It is the fourth-largest producer globally at roughly 8 percent of world output, anchored by Lynas Rare Earths' Mount Weld mine, the highest-grade rare earth deposit in the world. Lynas operates the largest separation facility outside China, in Malaysia, with additional processing under development in Texas and Western Australia. Australia demonstrates that reserve size matters less than ore quality, processing capability, and political stability.

Reserves vs. Production vs. Processing: A Comparison

The table illustrates the core asymmetry: Vietnam and Brazil together hold more reserves than China, yet their combined production is a rounding error next to Chinese output. The United States produces the second-most ore but refines little of it domestically. Reserve rankings are a leading indicator of future supply potential, not a description of today's supply security.

Why Reserve Data Is Uncertain and Often Misused

Several common mistakes distort public discussion of rare earth reserves. First, people conflate 'reserves' with 'resources.' Resources include material that is geologically identified but not economically viable; reserves are the subset that is. Headline-grabbing deposit announcements often cite resource figures that may never be mined. Second, reserve estimates are revised frequently and vary between sources; the USGS, national geological surveys, and company filings can differ by tens of percent for the same country. Third, 'rare earths' is a basket of 17 elements with wildly different market dynamics. Light rare earths like cerium and lanthanum are oversupplied and cheap, while heavy rare earths like dysprosium and terbium are scarce, expensive, and almost entirely refined in China. A country rich in light rare earths does not solve the heavy rare earth problem.

Fourth, reserve figures say nothing about time-to-market. Bringing a new rare earth mine from discovery to production typically takes 10-15 years or more, given permitting, environmental review, pilot processing, and offtake agreements. The 2025-2026 wave of government funding, including US Department of Defense equity stakes and price-floor mechanisms for MP Materials, EU Critical Raw Materials Act targets (10 percent domestic extraction, 40 percent processing by 2030), and Japanese and Korean offtake financing, is designed to compress that timeline, but geology and permitting set hard limits. Anyone citing reserve tables as proof of imminent supply diversification is misreading the data.

How AI Is Changing Exploration and Reserve Discovery

The reserve map is not static, and the biggest near-term changes may come from how new deposits are found. Traditional exploration is slow and expensive: only a small fraction of exploration targets ever become mines, and greenfield discovery costs have risen for decades. AI-powered exploration platforms are changing the economics. By training machine learning models on satellite imagery, geochemical surveys, geophysical data, and historical drill results, these systems can rank vast territories for rare earth prospectivity in weeks rather than years. The US Department of Energy has funded AI tools specifically to accelerate critical mineral discovery, and European startups in the geology-AI space have raised significant venture capital, including a €22 million round reported by EU-Startups for a Paris-based platform applying AI to transition-critical mineral discovery.

For rare earths specifically, AI exploration matters because the deposits are geochemically distinctive. Carbonatites, alkaline igneous complexes, and ion-adsorption clays each have recognizable signatures in multispectral satellite data and regional geochemistry, making them well-suited to machine learning classification. A platform that can screen an entire country's geology and flag high-probability carbonatite targets, for example, can compress the early-stage exploration funnel dramatically. This is where the next additions to the global reserve table will likely come from: not from reclassifying known deposits, but from AI-assisted discovery of carbonatite and clay-hosted systems in underexplored regions of Africa, South America, Canada, and Australia. For investors and policymakers, the practical implication is that today's reserve rankings will look conservative by 2030, and the countries that pair good geology with modern exploration technology will climb the table fastest.

Practical Steps for Tracking and Acting on Reserve Data

If you need reliable rare earth reserve intelligence, start with the USGS Mineral Commodity Summaries, published free each January, which provides the standard country-by-country reserve and production table. Cross-check against company technical reports (NI 43-101 or JORC filings) for deposit-level detail, and against the International Energy Agency's critical minerals reports for demand-side context. Track policy developments separately: China's export licensing regime, US Defense Production Act and DoD offtake actions, and the EU Critical Raw Materials Act milestones all change the effective value of reserves faster than geology does.

For companies and investors, the actionable sequence is straightforward. First, identify which rare earth elements your exposure actually depends on; heavy rare earth supply risk is far more acute than light. Second, evaluate projects on processing route and permitting jurisdiction, not just tonnage. Third, monitor AI-exploration platforms and their discovery announcements, because a credible new heavy rare earth clay discovery outside China can move project valuations and national reserve tables within a single reporting cycle. Fourth, remember the 10-15 year development lag: decisions made in 2026 about exploration funding and offtake will determine the 2035 supply map. The window for influencing that map is now, and the countries and companies that treat reserve data as a starting point for deeper analysis, rather than a final answer, will make better decisions than those that simply read the top line of the USGS table.