Table of Contents

The United States and NATO have various concerns about Russia’s large force of NSNWs, including the opacity surrounding those weapons’ locations. Consistent with Russian commitments under the 1991–1992 Presidential Nuclear Initiatives, Moscow has stated that the warheads for its NSNWs are “stored in centralized storage facilities,” which means that they are consolidated away from delivery systems (not that storage facilities are centrally located within Russia).1 While the United States has not publicly disputed this claim, the exact locations of Russia’s nonstrategic nuclear warheads remain unverified.2

The United States’ only NSNWs are gravity bombs, which are reportedly held at six air bases in Europe and various additional locations in the United States.3 The U.S. government has never confirmed these locations—not least to avoid drawing public attention to weapons whose presence their host states’ populations generally oppose—and influential Russian analysts have sometimes expressed concern that U.S. nuclear gravity bombs may be stored at other European locations.

Russia may believe that ambiguity surrounding the locations of its NSNWs enhances its ability to deter NATO. Ambiguity, however, also creates real risks. NATO states, for example, have long been concerned about the possible presence of nonstrategic nuclear warheads in the Russian enclave of Kaliningrad.4 It remains unclear, however, whether the recently renovated storage facilities there currently house such warheads or whether Russia wants the option to introduce them at short notice.5 When coupled with associated concerns that Russia might use NSNWs early in a conflict, the possible presence of nuclear warheads in the enclave is threatening and exacerbates tensions—potentially unnecessarily if, in fact, no warheads are present.

Moreover, in a conventional conflict, ambiguity could prove escalatory. If NATO wrongly believed that nuclear warheads were present in Kaliningrad, it might attack warhead storage facilities and potential dual-use delivery systems to forestall Russian nuclear use. Moscow might then conclude that such attacks represented the start of a broader conventional campaign against Kaliningrad, which is surrounded by NATO states and could be difficult to defend with exclusively nonnuclear means. As a result, Moscow might resort to the very nuclear use that NATO had sought to forestall.

These risks underscore the value of a cooperative effort to reduce uncertainty and potential misperceptions about the locations of nonstrategic nuclear warheads. Specifically, the absence of warheads at empty actual or suspected storage facilities could be verified through inspections. In this context, “empty” connotes the absence of all nuclear warheads, regardless of type. (Because of the difficulty of distinguishing strategic from nonstrategic warheads, it would be infeasible to inspect facilities at which strategic warheads were present but nonstrategic ones were not.) The goal here would be to enhance transparency around a limited number of pairs of facilities—say, two to five—of particular concern to NATO or Russia, not to create a comprehensive verification regime intended to cover every facility that has stored warheads or could do so.

Inspections of empty storage facilities would also facilitate progress toward the long-term goal of a more comprehensive management regime for NSNWs—an idea long supported by Washington in one form or another.6 Such a regime, however, would face various practical difficulties, as U.S. officials have acknowledged in calling for Russia and the United States to undertake joint experiments to overcome verification obstacles.7 The inspections proposed here would contribute to the development of a verification regime for a treaty, say, that prohibited NSNWs from being located anywhere except declared facilities, or more ambitiously, one that imposed a single limit on all warheads—strategic and nonstrategic, regardless of their deployment status (see chapter 9). This latter concept would require counting nondeployed warheads inside storage facilities. Inspections of empty facilities would advance this goal by creating joint understandings about, for example, how to prevent inspectors from learning classified information about storage facilities (such as their security features) and how to reach agreement on which areas of military bases should be open for inspections.

Solution Concept

Inspections of actual or suspected warhead storage facilities—even empty ones—would represent a new frontier for arms control but would build on experience from previous agreements. Russia and the United States would first need to negotiate a generic verification protocol. Then to implement the protocol, NATO and Russia would have to negotiate and select pairs of sites—each pair containing one facility on NATO territory (including the United States) and one on Russian territory. Sites would be selected on the basis of mutual consent from Russia, the United States, and, in the case of a NATO facility located outside of the United States, the host state. Establishing an iterative process, in which lessons from each round of inspections were used to improve the generic verification protocol, would be particularly useful.

NATO has previously considered a similar concept and, as a first step, attempted to draw up a list of former warhead storage facilities that Russia could inspect (presumably on the basis of reciprocity). However, this process stalled amid bureaucratic and legal difficulties, including those created by many such facilities now being privately owned. The process proposed here would be simpler. NATO would not have to determine a comprehensive list of sites for inspection. Instead, it would have to assess the feasibility of inspections only at those sites of interest to Russia.

Overview of an Inspection Regime to Verify the Absence of Nuclear Warheads at Selected Empty Actual or Suspected Warhead Storage Facilities

Russia and the United States should agree, on a politically binding basis, to reciprocal inspections of between two and five pairs of empty actual or suspected warhead storage facilities to demonstrate that they do not contain nuclear warheads.

They should then negotiate a generic verification protocol to verify the absence of nuclear warheads at an empty storage facility.

Russia and NATO should each propose, on the basis of its own intelligence information, candidate facilities that it wishes to inspect. From these lists, Russia and NATO should then jointly select one pair of facilities—one on NATO territory (including the United States) and one on Russian territory—for inspection. Facilities must be selected on the basis of mutual consent by Russia, the United States, and, for a facility located in a NATO state other than the United States, the host government.

In permitting a facility to be inspected, the host government would formally notify the inspecting state of the absence of nuclear warheads at the facility.

The United States and the host government would jointly facilitate all implementation activities for an inspection of a facility on NATO territory outside the United States.

Inspections to verify the absence of nuclear warheads should take place within sixty days of facility selection, following the exchange of baseline information and the negotiations over site-specific implementation arrangements.

After the first round of inspections, Russia and the United States should consult to discuss any problems that arose and refine the generic verification protocol accordingly. They should then repeat this process at one or more additional pairs of facilities.

There would be some risk of nuclear warheads’ being removed from a facility in the time between facility selection and inspection—though the inspecting state would certainly monitor the facility closely with national technical means (NTM). While this risk would not be acceptable in a more comprehensive treaty, it should be tolerable in a politically binding confidence-building measure. In fact, this problem could be solved in a treaty with a provision that permitted inspectors to notify the host state of the inspection site only after their arrival in country, as is standard practice.


Baseline information exchange. A prerequisite to on-site inspections would be the exchange of baseline information about the selected facilities and the negotiation of site-specific implementation details. A key challenge here would be reaching agreement on the boundaries for inspection activities. Inspections would be feasible only if Russia and the United States could agree where nuclear warheads might conceivably be stored in a facility—thus precluding the need to inspect highly sensitive areas, such as communication centers, and avoiding very disruptive inspections of, for example, barracks or offices.

Verification: Baseline Information Exchange

Within ten days of facility selection, Russia and the United States, which should act jointly with the host state in the event of an inspection on the territory of another NATO member, should provide the other with simplified site diagrams that show the boundaries of the military base or other site on which the selected facility is located, all fixed structures within that boundary, and the prospective boundaries for inspection activities. The states should then negotiate, and mutually consent to, any changes to the prospective inspection boundaries.

Within twenty-five days of facility selection, Russia and the United States should exchange the following detailed baseline information about the selected facilities:

  • A detailed inspection site diagram that conveys the agreed-upon inspection boundaries and the locations of any fixed structures within the inspection boundaries
  • Diagrams of each fixed structure’s interior, denoting all rooms (above- and below-grade) and interior and exterior access points
  • A list of each warhead storage container and each storage container for nonnuclear munitions located within the inspection boundaries, denoting the container type and location
  • The physical dimensions and a photograph of each type of warhead storage container located within the inspection boundaries
  • The physical dimensions and a photograph of each type of storage container for nonnuclear munitions located within the inspection boundaries

After the exchange of information, Russia and the United States should address any questions regarding the selected facilities and inspection procedures.

In implementing these provisions, the following definitions would apply:

  • “Fixed structure” means a unique structure that was previously used to store nuclear warheads, is designed to store nuclear warheads, or has an access point that is large enough for a warhead storage container to pass through.
  • “Room” means an interior subdivision within or underneath a fixed structure with an interior that has an access point that is large enough for a warhead storage container to pass through.8
  • “Storage container for nonnuclear munitions” means any weapon container, other than a warhead storage container, whose linear dimensions when measured at their widest points are equal to or exceed the dimensions of the test object.

On-site inspections. The data exchange would enable each state to plan and conduct its inspection. One complication is the potential presence of weapon storage containers—whether empty storage containers for nuclear warheads or containers used to store nonnuclear munitions. Additionally, nonnuclear munitions outside of containers and other sensitive equipment, such as components for security systems, may also be present. The inspection protocol seeks to balance the intrusive access needed for credible verification with the protection of classified information that is not germane to the task of verifying the absence of nuclear warheads.

The protocol gives the host state the right to shroud any objects it deems to be sensitive. Meanwhile, it gives inspectors access rights to any location where a warhead could be hidden—as determined by the use of a cylindrical test object, representing the smallest plausible warhead storage container. If any warhead storage containers are present, inspectors may look inside to verify the absence of warheads (doing so should not reveal classified information unless a warhead is actually present in contravention of the state’s declaration). Inspectors can also use radiation detection equipment to verify that shrouded objects and storage containers for nonnuclear weapons do not contain nuclear material. Such technology has been used successfully pursuant to various previous U.S.-Russian arms control agreements. Its application here could be somewhat more challenging because undeclared nuclear warheads could be more heavily shielded in storage than when deployed on delivery systems. On paper, this problem could be addressed by increasing the detection time—an issue that Russia and the United States could jointly consider while negotiating the generic verification protocol.

Verification: On-site Inspections

Inspections should occur within sixty days of site selection. The inspecting state should inform the host state of the inspection date at least forty-eight hours in advance of the inspection team’s arrival in country. All movement of vehicles and objects into or out of the inspection boundaries must cease twenty-four hours in advance of the inspection team’s arrival in country.

Preliminary Inspection Procedures

After the inspection team’s arrival at the facility, all vehicle traffic within the inspection boundaries must cease.

The host state’s escort team should notify the inspection team of any changes in the previously supplied facility information and provide updated inspection site diagrams (though the inspection boundaries must not be altered). The escort team should also note the locations of any inspectable vehicles situated within the boundaries. The inspection team should be permitted to designate up to two such vehicles for inspection.

The escort team should present the cylindrical test object to the inspection team. The inspection team should have the right to measure the object to ensure that it conforms to the agreed length and diameter dimensions.

If there are any warhead storage containers or storage containers for nonnuclear munitions at the facility, the inspection team should be permitted to view one of each type to verify their dimensions and to compare them to the photographs provided as part of the baseline information exchange.

Once preliminary inspection procedures are complete, any vehicles located within the inspection boundaries, other than those that have been designated for inspection, may resume movement.

Facility and Designated Vehicle Inspections

The duration of the facility inspection should be limited to twelve hours. During that period, for the sole purpose of verifying the absence of nuclear warheads, the inspection team should be given access first to any inspectable vehicles designed for inspection and then, in whatever order the team chooses, to any warhead storage containers, storage containers for nonnuclear munitions, and rooms that it selects. After each designated inspectable vehicle (if any) has been inspected, the host state may move it. The escort team should be permitted to accompany the inspection team at all times.

The host state should have the right to shroud, in advance of the inspection, any items other than warhead storage containers that it deems sensitive. The inspection team should have the right to request that the escort team open any warhead storage container to verify that it does not contain a nuclear warhead. The inspection team should also have the right to employ radiation detection equipment to confirm that any shrouded objects, storage containers for nonnuclear munitions, or other objects (other than warhead storage containers) do not contain nuclear material.

Undeclared Structures, Objects, and Vehicles

Should the inspection team observe, within the inspection boundaries, an undeclared fixed structure, warhead storage container, storage container for nonnuclear munitions, room, or inspectable vehicle, the escort team should be required to perform measurements to verify whether it is large enough to contain a nuclear warhead. If it is, inspectors must be provided with access. The duration of such access should not count toward the twelve-hour time limit for the inspection.

Access Point Characterization

An access point should be deemed large enough for a warhead storage container to pass through if the test object, in any orientation, can pass through it. The test object should be a cylinder with a length of 1 meter (3.3 feet) and a diameter of 0.5 meters (1.6 feet) and be constructed of lightweight materials.9

In implementing these provisions, the following definition would apply:

  • “Inspectable vehicle” means a vehicle that has an access point large enough for a warhead storage container to pass through.


Technical feasibility. In crafting a verification protocol, Russia and the United States could draw upon their considerable experience of conducting inspections at sensitive facilities and using radiation detection equipment for verification purposes. Moreover, they would not have to address the various complications associated with inspecting facilities at which nuclear warheads are present. Yet significant challenges—such as the need to agree on inspection boundaries—would remain. These challenges appear manageable, but the only way to know for sure would be for Russia and the United States to try to negotiate and implement the proposed agreement.

Political feasibility. The United States and its NATO allies have clear concerns about the locations of Russia’s nonstrategic nuclear warheads. And there is some evidence that these concerns are reciprocal.10 But even if Moscow does not currently have concerns about the locations of U.S. NSNWs in Europe, it could easily become concerned in the context of a reinvigorated arms race. Therefore, all states would benefit directly from enhanced transparency. This proposal would benefit the United States and NATO more than Russia, however, and so it would likely need to be paired with a measure of more interest to Russia—perhaps the confidence-building regime for Aegis Ashore launchers in Europe (see chapter 3).

At least three other political challenges could also arise. First, the proposed agreement may go too far for Russia and not far enough for the United States. The United States seeks a comprehensive regime for managing NSNWs. However, Moscow has repeatedly rejected this concept; indeed, it appears to be sticking to its long-standing position that U.S. nuclear weapons based in Europe must be returned to the United States and their storage infrastructure permanently dismantled before negotiations on limiting NSNWs can begin.11 This proposal could help to break the logjam. It would not require negotiations over limits on NSNWs but would build experience and confidence in inspecting warhead storage facilities, thus making progress toward the goal of a more comprehensive treaty.

Second, it could be challenging for Russia and the United States to select facilities for inspection. There would presumably be some empty actual or suspected warhead storage facilities at which an inspection request could not be accommodated; some are likely too sensitive, and others, such as former warhead storage facilities now in private hands, may present insurmountable bureaucratic and legal difficulties. Because facilities must be selected by mutual consent, the host state could always veto an inspection request. This veto power represents an important safeguard—though if it were used too often, the proposed agreement would likely fall apart amid reciprocal accusations of bad faith.

Finally, inspections in Europe could create particular legal, diplomatic, and political complications. While navigating these difficulties would not be entirely straightforward, they were successfully tackled in implementing the INF Treaty, which suggests that they should not be insurmountable.


1 Russian Federation, “Practical Steps of the Russian Federation Towards Nuclear Disarmament,” 2015 Review Conference of the Parties to the Treaty on the Non-Proliferation of Nuclear Weapons, New York, April 27–May 22, 2015, 20,

2 U.S. Department of State, “Adherence to and Compliance With Arms Control, Nonproliferation, and Disarmament Agreements and Commitments,” 12.

3 Kristensen and Korda, “United States Nuclear Weapons, 2020,” 56.

4 Hans M. Kristensen, “Non-Strategic Nuclear Weapons,” Special Report 3, Federation of American Scientists, May 2012, 70-75,

5 Hans M. Kristensen, “Russia Upgrades Nuclear Weapons Storage Site in Kaliningrad,” Strategic Security (blog), June 18, 2018,

6 U.S. Department of State, Office of the Spokesperson, “Secretary Blinken’s Call With Russian Foreign Minister Lavrov,” February 4, 2021,

7 U.S. Department of State, “Briefing With Ambassador Marshall Billingslea, U.S. Special Presidential Envoy for Arms Control and Lt. Gen. Thomas Bussiere, Deputy Commander of the U.S. Strategic Command,” August 18, 2020,

8 This definition would include, for example, underground weapon storage vaults.

9 These dimensions are illustrative. The dimensions of the test object should be based on intelligence information about the size of actual warhead storage containers. For a rough precedent for the use of a template, see Joseph P. Harahan, On-Site Inspections Under the INF Treaty: A History of the On-Site Inspection Agency and INF Treaty Implementation, 1988–1991 (Washington, DC: U.S. Department of Defense, On-Site Inspection Agency, 1993), 82,

10 Personal communication, Russian analyst, 2013.

11 Ministry of Foreign Affairs of the Russian Federation, “Deputy Foreign Minister Sergey Ryabkov’s Interview With the Newspaper Kommersant,” September 22, 2020, Ryabkov’s comments about negotiations over NSNWs were slightly ambiguous—hence the phrase “appears to be sticking.”