US wields nuclear test claim to press China on arms control

A disputed Chinese nuclear test has become the latest stress test for a global arms control system already coming apart.

At the Conference on Disarmament, the US accused China this month of carrying out a covert nuclear explosive test in June 2020. The allegation came just a day after the US–Russia nuclear arms control treaty, known as New START, expired and as the US presses for a broader new framework that includes China.

US Under Secretary of State for Arms Control Thomas DiNanno said China had conducted at least one “yield-producing” test and used “decoupling” to muffle seismic signals, arguing the activity violated test-ban commitments despite China’s declared moratorium.

Nuclear decoupling involves detonating a device in a large underground cavity to direct the explosion’s energy into compressing air, thereby reducing seismic signals. This makes nuclear tests seem smaller but doesn’t completely hide the explosion.

China rejected the US allegation. Its envoy, Shen Jian, dismissed the accusation as “false narratives” and said the US was fueling an arms race. At the same time, the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) said its monitoring network detected no explosion consistent with the US claim.

The legal backdrop complicates the dispute. Both the US and China signed the Comprehensive Nuclear-Test-Ban Treaty (CTBT) in 1996, but neither has ratified it, meaning the treaty has not entered into force and is not legally binding. Even so, both countries have observed voluntary moratoria on nuclear explosive testing since the late 1990s, helping sustain a global norm against testing.

The accusation also arrives amid a broader shift in US policy. In late 2025, US President Donald Trump ordered preparations to resume US nuclear testing “on an equal basis,” without elaborating what that means.

Trump has cited China and Russia’s alleged nuclear tests – again without clarifying whether that meant flight-testing nuclear-capable missiles or conducting explosive nuclear tests – as reasons for his order. He also said he has discussed nuclear arms control with Russia and wants to involve China.

If China did conduct a nuclear test in June 2020 and the US had indications at the time, the more revealing question is why the US chose to make the allegation public only now. In a Reuters report this month, Tomas Nagy notes that the US may have assessed that China would be unlikely to cooperate on arms control in the near term, prompting the US to make the accusation public for maximum political impact. But, if the US had compelling evidence in 2020, it is difficult to explain why the issue would have been muted for years.

That timing points to a broader policy calculation rather than a purely retrospective disclosure. It is plausible that the US is using the accusation to create political and strategic justification for renewed nuclear testing, regardless of whether China actually tested a device.

As Jack Burnham and Andrea Stricker argue in a Foundation for Defense of Democracies (FDD) article this month, the US charge should be read in the context of collapsing arms control and a wider rethinking of nuclear policy, not as an isolated factual claim.

Burnham and Stricker frame the accusation as signaling US readiness to expand or diversify its nuclear forces if arms control fails, suggesting the charge serves strategic and policy-shaping purposes beyond establishing what may or may not have happened in 2020.

Seen in that light, DiNanno’s statement looks less like a narrow compliance dispute and more like an effort to bring China into the existing arms control regime. China today is estimated to have around 600 nuclear warheads—roughly a tenth of the US or Russia’s stockpiles—and has little incentive to cap production while it is still moving from a minimal deterrence posture toward something closer to a first-strike-capable force comparable to the US and Russia.

From China’s perspective, joining a framework like New START before reaching that level would lock in permanent inferiority, unless the US and Russia were prepared to scale back their own arsenals dramatically. Framing the issue around alleged, unproven testing may therefore be less about physics than about political leverage.

At the same time, the technical balance also matters. Nuclear testing at the US Livermore laser fusion facility is almost certainly sufficient for testing any new warheads the US wants to develop. China is building a similar facility in Mianyang, Sichuan province, but it will take several years before it has comparable capability. Russia does not appear to have anything equivalent at present.

If those restraints are set aside, the practical consequences could be significant. David Sanger and William Broad report in a New York Times article this month that US officials are weighing reopening missile tubes on submarines, fielding new theater-range nuclear weapons, completing modernization programs and preparing to conduct underground nuclear tests for the first time since 1992.

There is, however, a real risk that the current cycle of accusation and signaling could backfire. Doreen Horschig argues in an April 2025 report for the French Institute of International Relations (IFRI) that accusations of covert testing and threats to resume testing are increasingly being used as strategic tools.

She notes that a strong post–Cold War norm against nuclear explosive testing, reinforced by CTBTO institutions and voluntary moratoria, made testing politically costly and rare. As arms control weakens, Horschig asserts, states—including the US—are using allegations, threats and readiness postures to signal resolve, shape negotiations and gain leverage over rivals.

Linking policy shifts to others’ alleged violations risks eroding the testing taboo, increasing mistrust and encouraging reciprocal action that could unravel decades of restraint.

Daryl Kimball and Xiaodon Liang make a similar case in an October 2025 statement for the Arms Control Association (ACA). They argue that making accusations about nuclear testing without shared, verifiable evidence is dangerous and counterproductive.

In their view, US claims that Russia and China conducted very low-yield tests are “unsubstantiated and highly debatable,” and such concerns are better addressed through CTBT ratification, inspections and confidence-building measures rather than political threats.

They warn that reckless claims and talk of renewed testing could trigger a chain reaction of nuclear testing, weaken the nonproliferation regime and harm US security, undermining both US credibility and arms control norms by replacing verification with accusation.

There are also signs that transparency, rather than accusation, still has practical value. Axios reported this month that the US and Russia reached an understanding in Abu Dhabi to continue observing key New START transparency practices even after the treaty’s formal expiration, preserving data exchanges and predictability despite the collapse of the legal framework.

Ultimately, the key to avoiding nuclear war is not weapons numbers or explosive testing but transparency and mutual predictability. In that context, even small nuclear tests that are mutually disclosed and verifiable could be stabilizing, while efforts to drag China into a narrative of secret violations risk weakening the very transparency that has done the most to keep nuclear rivalry from escalating into catastrophe.