GAO Weapon Systems Annual Assessment 2025 (GAO-25-107569)
The most detailed open dataset on US defense cost growth: GAO's 23rd annual review of the Department of Defense's costliest weapon programmes. Here is what GAO-25-107569 found, the portfolio it covers, and how to cite it.
What the report found
GAO-25-107569 assessed 106 major US weapon programmes representing nearly $2.4 trillion in planned investment. Combined cost estimates for the 30 continuing programmes also in the prior year's report rose $49.3 billion in a single year, with the Sentinel ICBM alone driving over $36 billion (73%) of that increase. The average time to deliver even an initial capability rose 18 months, to almost 12 years from a programme's start.
Source: US Government Accountability Office (2025). Weapon Systems Annual Assessment, GAO-25-107569, 11 June 2025.
What the assessment covers
Each year GAO reviews the Department of Defense's most expensive acquisition efforts and reports cost, schedule, and performance against their approved baselines. The 2025 edition covered 106 programmes: major defense acquisition programmes (MDAPs) plus Middle Tier of Acquisition (MTA) programmes, together representing nearly $2.4 trillion in planned investment. It is the closest thing to an audited, programme-by-programme public ledger of cost growth that exists for any government or industry.
The report also assessed 20 of the costliest MTA programmes, which carry about $44.5 billion in planned investment and are intended to deliver in two to five years. GAO's recurring theme is that programmes which secure key knowledge, such as technology maturity, before committing to production tend to post better cost and schedule outcomes than those that do not.
Why the cost estimates keep rising
The $49.3 billion one-year increase is concentrated, not spread evenly. A single programme, the Sentinel intercontinental ballistic missile, accounted for over $36 billion, or 73%, of the rise, after it breached a critical Nunn-McCurdy cost threshold and was restructured. That concentration is the same fat-tail pattern seen across civilian megaprojects: a handful of programmes drive most of the portfolio's overrun.
On schedule, GAO found the expected time to deliver even an initial capability climbed 18 months in a single year, to almost 12 years from a programme's start. GAO also reported that most of the 39 programmes using a modern software development approach still delivered working software to users more slowly than industry Agile practice recommends, which is the basis for the report's call to structure newer programmes for speed and iterative delivery.
Where it sits among the large-sample overrun studies
The GAO assessment is the authoritative source for US defense cost growth specifically. It sits alongside the cross-sector datasets that anchor the rest of this site.
| Study | Sample | Headline finding |
|---|---|---|
| GAO Weapon Systems Assessment (2025) | 106 major DOD programmes | $49.3bn one-year cost-estimate rise; ~12 yrs to initial capability |
| Flyvbjerg project database (2023) | 16,000+ projects (all sectors) | Defense mean overrun high; only 0.5% deliver on budget, time and benefits |
| Standish CHAOS (2020) | 50,000+ project records | 31% successful, 50% challenged, 19% failed |
| PMI Pulse of the Profession | Global PM survey | Cross-sector: ~43% of projects exceed their original budget |
Cite the report number, not just “the GAO study”
GAO publishes a new Weapon Systems Annual Assessment every year, and the headline figures change with each edition. The 2025 edition is GAO-25-107569, published 11 June 2025, and it is the 23rd in the series. When quoting a figure such as the $49.3 billion one-year increase, cite the specific report number and year so the number stays traceable to the edition it came from.
Programme-level overrun figures for individual systems (F-35, Sentinel, Columbia-class, Ford-class) are collected on the defense projects page, which draws on this and prior GAO assessments.
How to cite
Primary source: gao.gov/products/gao-25-107569
The full report and a one-page Highlights summary are free on gao.gov. The Highlights PDF carries the headline figures; the GAO Weapons Acquisition portal lists every edition back to FY01.
Frequently Asked Questions
What is GAO-25-107569?
GAO-25-107569 is the US Government Accountability Office's 2025 Weapon Systems Annual Assessment, published 11 June 2025. It is the 23rd edition of GAO's annual review of the Department of Defense's costliest weapon programmes, and it assessed 106 programmes representing nearly $2.4 trillion in planned investment.
How much did US weapon programme cost estimates increase in 2025?
Combined cost estimates for the 30 continuing MDAPs also in the prior year's report rose $49.3 billion in a single year. The Sentinel ICBM programme alone accounted for over $36 billion, or 73%, of that increase, after a critical Nunn-McCurdy breach and restructuring.
How long do major US weapon programmes take to deliver?
GAO found the expected time to deliver even an initial capability rose 18 months in 2025, to almost 12 years from a programme's start. That average includes MDAPs that began as Middle Tier of Acquisition programmes, which are meant to deliver in two to five years.
Where can I read the report?
It is free on gao.gov at gao.gov/products/gao-25-107569, with a full PDF and a one-page Highlights summary. GAO publishes a new edition each year, so cite the specific report number and year.