Where Does North Korea’s Nuclear Program Stand in 2026?

Nearly two decades after international inspectors were expelled from North Korea, satellite imagery and open-source analysis remain the only windows the outside world has into Pyongyang’s nuclear program. What that window shows in late 2026 is a program that is visibly expanding, on multiple fronts at once.

On August 28, the International Atomic Energy Agency (IAEA) issued a report to its Board of Governors documenting the construction of a second uranium enrichment facility at North Korea’s main Yongbyon nuclear complex. The new facility could house up to 28 centrifuge cascades, alongside an already-operating expanded annex at the Kangson enrichment site. The agency identified the new two-story building by comparing satellite images taken during its construction with state media photographs of a June 2026 visit by Kim Jong Un, during which centrifuges and cascades resembling those at known enrichment sites were visible.

IAEA Director General Rafael Grossi told the Board that the continued operation of the Kangson and Yongbyon facilities, together with the reported further expansion of weapons-grade nuclear material production, was “a cause for serious concern.” The IAEA has not had inspectors on the ground in North Korea since 2009 and now builds its assessments entirely from commercial satellite imagery and open-source material, including North Korean state media itself.

The picture built up over 2026 by independent monitors is consistent with the IAEA’s findings. Analysts tracking Yongbyon have documented near-continuous water discharge from the 5-megawatt-electric (MWe) reactor, North Korea’s original plutonium producer, consistent with it running through what is now believed to be its seventh irradiation cycle since October 2024, alongside periodic steam activity at the Radiochemical Laboratory suggesting reprocessing campaigns. A separate site review by Daily NK reached similar conclusions, also noting that the long-delayed Experimental Light Water Reactor (ELWR) is showing renewed discharge activity after a shutdown period, and that demolition of the old, unfinished 50 MWe reactor at Yongbyon continues in parallel with the new construction.

A review by 38 North covering 2025 activity framed this as a direct response to political direction: Kim Jong Un had described that year as decisive for “bolstering up the nuclear forces,” and the satellite record showed the 5 MWe reactor running steadily since January 2025. 38 North also flagged pre-operational testing of the ELWR, a probable summer reprocessing campaign, and construction of what are assessed to be new enrichment and waste-storage buildings.

North Korea’s declared and suspected enrichment infrastructure now spans at least two locations at Yongbyon (the original uranium enrichment plant and the new building identified in August) plus the Kangson facility outside Pyongyang, whose 2024 annex expansion appears to have become operational around January 2026. Because clandestine centrifuge halls are comparatively easy to conceal, nonproliferation researchers have long argued that the more reliable way to bound North Korea’s enrichment capacity is to track its uranium supply rather than guess at hidden buildings. In 2021, an academic study using satellite-based land-use analysis around the Pyongsan uranium mill concluded that North Korea’s tailings output implies it could produce substantially more milled uranium than it is believed to currently be enriching, meaning supply is not the constraint on how fast the program could grow.

Unlike uranium enrichment, North Korea’s plutonium supply chain runs through a small number of well-mapped facilities. It starts at the Pyongsan Uranium Concentrate Plant, roughly 45 kilometers from the DMZ, which remains the country’s only confirmed source of yellowcake. Uranium ore is processed there into oxide concentrate before being shipped to Yongbyon for conversion into reactor fuel. That fuel feeds the 5 MWe reactor, whose spent fuel rods are then reprocessed at Yongbyon’s Radiochemical Laboratory to extract weapons-grade plutonium. This lab is believed to be responsible for the totality of North Korea’s plutonium stockpile to date, estimated by outside experts to support somewhere in the range of dozens of warheads.

The ELWR adds a second, more ambiguous track. Rated at 100 megawatts-thermal, it began sustained operation in late 2023 after a decade of delay. The ELWR has since gone through at least one documented shutdown-and-restart cycle, with Grossi telling the IAEA board of direction in November 2025 that it had “likely been shut down” in early August of that year, after months of stable running. As noted above, the ELWR is believed to have restarted once again.

A light-water reactor of this design is not optimized for weapons-grade plutonium production the way the graphite-moderated 5 MWe unit is, and North Korea has publicly framed it as an electricity source. Even so, should it reach stable, full-power operation, outside analysts assess it would function as a supplementary plutonium source layered on top of the existing reactor, meaningfully increasing Yongbyon’s total fissile material output.

The reactor most often described in older reporting as North Korea’s “second 200-megawatt reactor” is not at Yongbyon at all, but roughly 20 kilometers away at Taechon. Construction of this larger, 200 MWe graphite-moderated, gas-cooled unit began in the late 1980s as the planned full-scale successor to the 5 MWe design but was frozen under the 1994 U.S.-North Korea Agreed Framework, at which point North Korea had manufactured none of the reactor’s required graphite blocks). The most recent detailed satellite review of the site, by 38 North in 2019, found administrative and support-building upgrades but no changes to the derelict reactor structure itself or its generator hall. The analysts concluded that renewed construction was less likely than signaling or preparation for eventual inspections. No subsequent commercial satellite analysis has documented a resumption of reactor construction at Taechon; as of 2026 it remains, by all available open-source evidence, an abandoned shell.

The reason Taechon still draws attention despite decades of dormancy is the scale of what it was designed to do. At full design output, a 200 MWe graphite-gas reactor of this type could produce on the order of 220 kilograms of weapons-grade plutonium per year – enough, by some open-source estimates, for 25 to 40 additional nuclear weapons annually, dwarfing the output of the 5 MWe reactor that has anchored North Korea’s plutonium stockpile for four decades. 

In practice, most nonproliferation analysts consider a Taechon restart unlikely in the near term: the design is nearly 40 years old, the site has been exposed to the elements for three decades, and North Korea’s more recent investment has gone into the ELWR and centrifuge halls rather than graphite-gas technology. But the reactor’s unfinished shell, still standing, is a reminder that North Korea’s ambition for plutonium production was never defined by its current 5 MWe reactor. 

There are three main takeaways from open source analysis of North Korea’s nuclear program in 2026. First, North Korea is continuing a steady, uninterrupted plutonium cycle at the original 5 MWe reactor. Second, its light-water reactor has started on-again, off-again operation, which could add a second plutonium stream if it stabilizes. Finally, its uranium enrichment complex has now visibly doubled its footprint at Yongbyon while continuing to expand at Kangson. 

The fact that North Korea did all this without meaningfully touching the Taechon project tells us that Pyongyang has found it more efficient to build new capacity rather than revive old designs. That may be the most consequential inference to be drawn from Taechon.