Highlights

• U.S. rare earth production surged from 4,300 to 8,900 metric tons in 2025, but import reliance rose to 67% despite domestic gains, signaling continued fragility.
• Heavy rare earths like dysprosium and terbium remain 100% import-dependent with no commercial-scale U.S. production, creating a strategic bottleneck for defense and EV applications.
• Record concentrate production of 51,000 metric tons masks the real challenge: downstream separation, metal-making, and magnet manufacturing capacity remain critically underdeveloped.

---

The U.S. rare earth story is finally moving in the right direction, and the latest U.S. Geological Survey (USGS) Mineral Commodity Summaries 2026 data (opens in a new tab) reinforce that. But the same tables also deliver a sobering message for investors and policymakers: the U.S. is building capacity—not yet command. The headline numbers look encouraging, yet the most strategic segments of the supply chain remain thin, import-exposed, and vulnerable to shocks. Note American treasure trove MP Materials is producing the vast bulk of the output as of the end of 2025.

What the USGS Numbers Say—and What They Don’t

USGS reports a sharp jump in U.S. production of rare-earth compounds and metals (expressed in rare-earth oxide equivalent) from 4,300 metric tons in 2024 to an estimated 8,900 metric tons in 2025. That is meaningful progress. It reflects years of capital, permitting, and operational learning, finally showing up in national statistics.

But investors should avoid a common translation error: “compounds and metals (REO equivalent)” does not automatically mean full-spectrum, separated, market-ready oxides across the board. It can include mixed or intermediate chemical forms reported as REO-equivalent for consistency. Treating the figure as proof of complete refining independence overstates what the data can support.

Import Reliance Fell—Then Rose Again

Another misunderstood talking point is “import reliance was cut in half.”  Directionally, yes: net import reliance fell from over 90% in 2023 to 53% in 2024, then rose to 67% in 2025, even as domestic output increased. That reversal matters. It suggests the system is still fragile, dependent on trade flows, and not yet structurally de-risked.

USGS also flags a major blind spot: rare earths embedded in imported finished goods—motors, magnets, electronics—can make headline import metrics look safer than real exposure.

The Hard Wall: Heavy Rare Earths

Here is the strategic cliff: heavy rare earths remain 100% net import reliant. USGS indicates that while minerals containing heavy rare-earth elements may be mined domestically, there was no sustained commercial-scale production of heavy rare-earth compounds or metals in 2025. That’s the choke point. Dysprosium and terbium are essential for high-coercivity magnets used in defense systems, drones, EV drivetrains, and industrial motors. Progress on light rare earths does not substitute for this gap.

Concentrate Records Aren’t the Finish Line

The USGS also reports a record mineral concentrate production of 51,000 metric tons of REO in 2025 (up from 45,500 in 2024). That’s real momentum—upstream. But strategic leverage comes downstream: separation, metal-making, alloying, magnet qualification, and manufacturing at scale. Those layers remain the U.S. bottleneck—especially for heavies.

A Reality Check from the Ore Body

In background discussions, one major U.S. producer, our American treasure trove MP Materials,  has emphasized a blunt truth: the ore body drives the mix. Roughly ~80%+ of many concentrates can be cerium and lanthanum—high-volume but low-value products in persistent oversupply—while NdPr is the economic engine. MP Materials sells NdPr oxide (not separated Nd and Pr), because most modern magnet recipes accept NdPr oxide and the natural Nd:Pr ratio typically fits market specs. On heavies, as _Rare Earth Exchanges_™ has pointed out, the company’s SEG+ stream includes ~4% dysprosium and terbium on a total rare earth oxide basis—small by percentage, meaningful by absolute volume when paired with high head grade and third-party feedstocks.  See the company’s literature (opens in a new tab).

REEx Take

The USGS data support optimism—but only disciplined optimism. The U.S. is building a foundation it did not have five years ago. But use of the word “de-risked” is still premature. Embedded imports mask real exposure. Heavy rare earths remain the strategic cliff. And downstream capability—not concentrate tonnage—will decide whether America’s rare earth comeback becomes durable.

See the latest USGS report (opens in a new tab).