How Much Would Resuming U.S. Nuclear Weapons Testing Cost?
Published on | Prices Last Reviewed for Freshness: November 2025
Written by Alec Pow - Economic & Pricing Investigator | Content Reviewed by CFA Alexander Popinker
CFA Educational content; not financial advice. Prices are estimates; confirm current rates, fees, taxes, and terms with providers or official sources.
Per-shot cost is $132–$146M (2024$). Staying “warm” runs $15–$30M/year. Don’t confuse that with the $946B decade-long modernization bill.
The U.S. hasn’t detonated a nuclear device since 1992. That may change: on October 30, 2025, Reuters reported that President Trump told the Pentagon to immediately prepare for renewed nuclear testing. The same week, the Kremlin announced a Poseidon trial.
The U.S. has observed a voluntary test moratorium since 1992; if testing returns, the likely venue is Nevada National Security Site (NNSS).
Think in ledgers: a test is a billable engineering project; “get ready” is a monthly meter; modernization is a decade program; RECA is social liability. Mixing them muddies the number.
- White House direction to prepare for testing: Reuters, Oct 30, 2025
- Russia’s Poseidon claim, used as political backdrop: Reuters, Oct 29, 2025
- Ten-year U.S. nuclear-forces estimate (often misread as “testing cost”): Reuters summary of CBO, Apr 2025
Myths & where they miss
Jump to sections
These are real talking points circulating in headlines, social posts, and threads. For each, we show where it appeared, why the reasoning breaks, and the grounded number you should use instead.
| Myth (as posted) | Where it spread | Why it misses (the receipts) |
|---|---|---|
| “Testing will cost nearly a trillion dollars.” Some viral posts lift a giant number and label it the ‘testing budget.’ | Headlines on long-horizon modernization (e.g., The Guardian, 2017) and the current 10-year CBO outlook (CBO, 2025). | Those are enterprise-wide costs (operations, delivery systems, facilities, command/control), not the price of a single underground shot. The per-test engineering bill sits around $132–$146M (2024$), per Sandia/FAS. Use modernization figures for decade-scale planning, and the per-test band for a discrete invoice. See: FAS summary. |
| “That $1.8B Nevada project is what a test costs.” Screenshots of a lab price tag get framed as ‘the cost of one test.’ | AP News on “Scorpius”, an underground diagnostics accelerator. | Scorpius is a multi-year facility for subcritical science and diagnostics, a capital project, not an explosive test. Conflating a lab build with a one-off underground shot is apples/oranges. A single test falls in the mid-nine-figure range; a facility can run near or above a billion because it’s built once and used for years. |
| “The U.S. can test immediately now.” Threads imply a same-week detonation after a political directive. | Coverage of an Oct. 30 directive to “prepare” (Reuters) gets paraphrased as “test now.” | “Prepare” is not “detonate.” Environmental review, legal process, security, and technical readiness take time and money. Historically, moving to a faster posture carried $83M over three years to reach ~18-month readiness from a cooler baseline, with an annual carry in the $25–$30M zone. Source: CRS RL32130 (PDF). |
| “Nuclear testing is illegal for the U.S. under the CTBT.” Posts treat a voluntary moratorium as a binding ban. | General discourse around the CTBT; see explainers and news (e.g., AP coverage) and a CRS note (CRS IF11662). | The U.S. has observed a testing moratorium since 1992 and signed the CTBT, but has not ratified it. That makes resumed testing a political/legal fight — not an automatic criminal act under U.S. treaty obligations. Costwise, the key is that “no test” still has a price if you hold a warm posture. |
| “Those $2.5–$2.6B U1a upgrades show what a test costs now.” Graphics equate an underground complex upgrade with a per-shot bill. | Government oversight reports on U1a instrumentation & related work (e.g., GAO-23-105714). | U1a is an infrastructure program supporting subcritical experiments and diagnostics, multi-year capital + integration. A single underground nuclear test is a discrete project with drilling, stemming, containment, diagnostics, and immediate monitoring, priced in the $132–$146M band (2024$). Don’t turn capex into a per-shot rate. |
| “Keeping teams warm is basically free.” Comments suggest costs only start when a device is fired. | Paraphrases under stories about “getting ready to test,” and budget threads lacking line-item context (e.g., Reuters + CRS RL32130). | Readiness burns cash: staffing posts, keeping instruments alive, retaining counsel, and holding contractors all carry a meter historically around $15–$30M/year (~$1.8–$3.0M/month) even with zero shots. That’s why delays show up as real dollars. |
Rule of thumb for readers: if the number describes a decade-long enterprise or a multi-year facility, it’s not a per-test price.
Media Accuracy Scoreboard
Real headlines readers confuse with a per-test price, and how to anchor them correctly.
| Outlet & link | What the story says | Why readers misread it | The correct anchor |
|---|---|---|---|
| AP News (2023): Scorpius | Underground particle accelerator project pegged near $1.8B. | A big underground number looks like a “test cost.”It’s diagnostics infrastructure. | Per-test is ~$132–$146M (Sandia-derived). Facility builds sit on separate capital lines. |
| Reuters (2025): CBO $946B | Ten-year enterprise-wide nuclear forces bill totals $946B. | Some readers lift the number as “the cost of testing.” It’s the whole nuclear enterprise, not one detonation. | Treat as modernization/operations context; do not use it as a per-test proxy. |
| PBS (2017): $1.2T over 30 yrs | Older CBO horizon: $1.2T modernization over ~30 years. | Longevity and scale can be mistaken as “testing costs,” especially in social posts. | A single underground test is a mid-nine-figure engineering project, not a trillion-dollar program. |
These are real, citable figures; they’re just different ledgers. When you see a big number, ask: “Is this a test invoice, a monthly readiness meter, or a multi-year enterprise plan?”
At the midpoint, one test (~$139,000,000) works out to about $1.06 per U.S. household. Rule of thumb: about a dollar per household.
The price for one underground test
A Sandia-derived estimate brought forward by the Federation of American Scientists places a single underground test at $132–$146 million (2024 dollars). The bundle covers site prep and drilling, stemming and containment, diagnostics and instrumentation, operations and security, and immediate post-shot monitoring. See: FAS summary of Sandia work.
For unit-price context on the device itself (separate from test logistics), see our breakdown of how much a nuclear bomb costs, including materials, design, and production factors.
The warm meter
“Get ready” is not free. A warm posture keeps crews current, instruments alive, posts staffed, and counsel retained. Historical midpoints imply $1.8–$3.0 million per month (~$60,000–$98,000 per day). Arithmetic: a $26M yearly posture ≈ $2.17M/month, with retention and legal premia pushing toward the top of the range. Source context: CRS RL32130 (PDF).
Who gets paid
Money flows through the Nevada National Security Site (NNSS) via a management-and-operations (M&O) prime and a web of lab agreements and subcontracts. Below are named entities historically tied to NNSS work and what checks they would plausibly collect in any return-to-testing scenario.
| Line of work | Examples of vendors / entities | What they’re paid for | Source |
|---|---|---|---|
| Site management & integration (prime) | Mission Support & Test Services (MSTS), a JV of Honeywell, Jacobs, and HII | Runs NNSS under NNSA’s M&O contract; integrates drilling, diagnostics, security, safety, and logistics; lets/oversees many subcontracts. | CRS IF10640 (2024) | Award coverage |
| Historical operators (context) | Bechtel Nevada (1995–2006) → National Security Technologies (NSTec) (2006–2017) | Earlier primes that managed the site pre-MSTS; useful for tracing legacy subcontractor ecosystems and capabilities. | DOE IG (2006) | NNSS (about) |
| National labs & diagnostics | Sandia, Los Alamos, Lawrence Livermore (program leads for subcritical science, timing, sensors) | Designs and fields diagnostics, timing & containment instrumentation; plans shots/experiments; analyzes data under Stockpile Stewardship. | FAS (Sandia analysis) | NNSS Area 27 (context) |
| Protective force & security | SOC (protective force, post-2018); previously Centerra (pre-2018) | Armed security, access control, perimeter posts, convoy/critical-asset protection during high-risk windows. | GAO-21-409 (2021) |
| Environmental & radiological monitoring | Navarro Research & Engineering (environmental services at NNSS) | Sampling plans, dose modeling, groundwater/air monitoring before & after events; public reporting and compliance. | DOE/CPAR (Navarro) | Contract news |
| Drilling, stemming & ground support | Let via MSTS subcontracts to specialty civil/mining firms (varies by task order) | Boreholes, instrumentation shafts, stemming plugs, ground support/repair, ventilation & access in underground drifts. | CRS IF10640 (MSTS prime) |
Notes: Specific subcontract awards are competed and may change by project phase. “Who gets paid” ultimately hinges on task orders issued under the MSTS M&O and related lab/agency agreements.
The MSTS owners (Honeywell, Jacobs, HII) are large federal contractors with active corporate PACs that donate to candidates in both parties. Sector-wide, defense-industry giving in recent cycles leaned modestly Republican but remained bipartisan overall. Reference: Taxpayers for Common Sense (2017–2022 trend).
The secrecy surcharge
Classified work shrinks the bidder pool. Only companies with cleared facilities, cleared staff, and compliance programs can compete, which naturally reduces price pressure. Those compliance costs (clearances, secure IT, compartmented areas, audits) get baked into overhead and show up on the invoice.
Secrecy also limits price discovery. Redacted contracts and restricted statements of work make it hard for the public (and sometimes competing vendors) to benchmark unit prices. Oversight bodies have repeatedly flagged cost growth and program-control gaps in weapons activities, signals that hidden work often carries hidden premiums. See, for example, the GAO management review (Sept 2025).
- Fewer bidders → higher unit prices: Only cleared firms can bid; less competition usually means less aggressive pricing.
- Compliance overhead: Facility clearances, secure networks, specialized logistics, and classification controls add fixed costs to every project.
- Limited price transparency: Redactions and need-to-know rules make it hard to benchmark or negotiate “best available” rates.
- Schedule risk premiums: Classified moves (convoys, access windows) and injunction risk lead vendors to price in pauses and overtime.
Translation: secrecy adds line items (compliance, security, access windows) that raise the total.
The Standing Cost of Staying Warm
Readiness isn’t free. Historically, when “test readiness” was explicitly funded, analysts recorded a carry around $25–$30 million per year (≈ $2.1–$2.5M/month). Earlier work also priced $83 million over three years to accelerate toward an 18-month posture from a slower baseline. Today, readiness activities tend to sit inside broader NNSA accounts rather than a single line, but the meter still runs. A reasonable midpoint illustration is $26M/year (historical, not current law). Source: CRS RL32130 (PDF).
At a midpoint of ~$2.17M/month for readiness, a three-month legal pause alone burns ~$6.5M. Add modest legal/retention premia and you approach $7–$9M, with no test conducted.
Setup Realities (not a single-shot bill)
The underground diagnostics backbone is being rebuilt for subcritical science and future experimentation. These are capital projects, not per-test invoices, but they frame the modern environment and timelines:
- UCEP at NNSS (~$876M): A multi-year underground experimental facility after independent review/scope refinement (ExchangeMonitor).
- U1a complex upgrades (~$2.5–$2.6B): Expanded instrumentation and access for subcritical experiments (GAO-23-105714).
- Ongoing NNSS milestones: Phased work and access improvements “a thousand feet below ground”.
These big numbers are infrastructure. They support science and any future testing, but they are not the price of a single shot.
Two Ledgers, One Debate
Keep two columns in mind. One column is the direct test (what engineers and site managers can cost out). The other is exposure (policy, legal, and social costs that may move independently of any shot).
| Direct test ledger | Exposure ledger |
|---|---|
| Drilling/boring and stemming; containment engineering; diagnostics & timing; field ops & security; first-pass environmental monitoring. | Environmental reviews, court challenges, diplomacy-driven delays, and compensatory programs such as RECA. |
| This can be estimated like any discrete engineering project. | Totals depend on policy choices and case outcomes, not strictly on “one test.” |
Since 1990, the Radiation Exposure Compensation Act has paid $2.6B+ to 41,000+ claimants; lawmakers discussed expansions in 2025. Keep RECA separate from a per-test invoice: CRS brief.
Direct test costs plus a simple readiness carry. Excludes RECA exposure, big facility projects, and diplomacy-driven delays.
| Scenario | Direct test count | Per-test midpoint | Direct subtotal | Readiness / yr | Window | Readiness subtotal | Illustrative total |
|---|---|---|---|---|---|---|---|
| Readiness only | 0 | n/a | $0 | $26,000,000 | 1 year | $26,000,000 | $26,000,000 |
| One-shot campaign | 1 | $139,000,000 | $139,000,000 | $26,000,000 | 1 year | $26,000,000 | $165,000,000 |
| Confidence series | 3 | $139,000,000 | $417,000,000 | $26,000,000 | 2 years | $52,000,000 | $469,000,000 |
| Prep then pause | 0 | n/a | $0 | $26,000,000 | 2 years | $52,000,000 | $52,000,000 |
| Two shots + extended monitoring | 2 | $139,000,000 | $278,000,000 | $26,000,000 | 2 years | $52,000,000 | $330,000,000 |
Midpoints use the $132–$146M per-test band and a $26M/year readiness carry. Swap in agency updates as they publish.
| NAICS / Line | Amount |
|---|---|
| 237990 Drilling & stemming | $48,000,000 |
| 541330 Containment engineering | $22,000,000 |
| 334515 Telemetry & sensors | $38,000,000 |
| 561612 Security & operations | $21,000,000 |
| Environmental monitoring | $10,000,000 |
| 562910 Post-shot sampling | $5,000,000 |
| Direct test subtotal | $144,000,000 |
| Warm posture, one year | $26,000,000 |
| Illustrative total | $170,000,000 |
Line items mirror the Sandia-derived bundle; your newsroom can re-map categories as agencies publish finer detail.
People and protection don’t pause for free; injunctions and reviews extend the meter.
| Monthly item | Low case | High case |
|---|---|---|
| Readiness burn incl. security | $1,400,000 | $2,100,000 |
| Contractor retention premiums | $250,000 | $500,000 |
| External legal & expert support | $150,000 | $350,000 |
| Monthly subtotal | $1,800,000 | $2,950,000 |
| Six-month subtotal | $10,800,000 | $17,700,000 |
A half-year delay can quietly add a low eight-figure bill before any stemming plug is poured or cable is calibrated.
The modernization backdrop
The 10-year price tag for U.S. nuclear forces is large, but not the same as a test invoice. The Congressional Budget Office places it at $946B across 2025–2034 for operations, delivery systems, facilities, command and control, and an overrun allowance. See: CBO (2025) or Reuters summary.
The enterprise-wide modernization bill also covers delivery systems. For a line-item view of the land-based leg, here’s how much an intercontinental ballistic missile costs from acquisition through sustainment.
Likewise, the sea-based deterrent is a major cost driver; see how much a nuclear submarine costs to build and operate, and why SSBN programs dominate capital outlays.
Answers to Common Questions
Where would a U.S. explosive test most likely occur?
At NNSS, which maintains subcritical and diagnostic capabilities and historically hosted U.S. underground shots.
How fast could a test occur under current practice?
Even with a directive, legal and environmental processes are lengthy; past readiness postures assumed months to accelerate.
Would testing reduce the modernization bill?
Not directly. A test may inform design or certification choices, but it does not erase the larger enterprise plan.
How should readers treat RECA in this discussion?
As separate exposure. The program has paid $2.6B+ and expansion was debated in 2025; it is not a per-test line item.
What single number should editors lift for a headline?
$132–$146 million per underground test (2024 dollars), with a warm posture around $15–$30M/year if sustained.
Bullet summary
- Per underground test: $132–$146M (2024$); midpoint ~$139M.
- Warm posture: ~$1.8–$3.0M/month (~$15–$30M/year), even at zero tests.
- Diagnostics & facilities (e.g., Scorpius, U1a/UCEP): substantial but separate from a per-test invoice.
- Nuclear-forces modernization (enterprise-wide): $946B over 2025–2034 (different ledger).
- RECA to date: $2.6B+ paid; expansion debated in 2025 (separate exposure).
Named sources used
Reuters reporting (Oct 2025); Federation of American Scientists (Sandia analysis, 2024); Congressional Research Service RL32130 & In-Focus notes on testing; Associated Press on Scorpius (2023); Congressional Budget Office (Apr 2025) and Reuters summary; NNSS news releases; GAO oversight (2025).
Leave a Reply
Want to join the discussion?Feel free to contribute!
People's Price
No prices given by community members Share your price estimate
How we calculate
We include approved comments that share a price. Extremely low/high outliers may be trimmed automatically to provide more accurate averages.