Cost estimating in DoD, explained

A program cost estimate is the Department of Defense (DoD) forecast of what a weapon or information system will cost, and it underpins budgets, baselines, and milestone decisions. Two guides set the practice. The Government Accountability Office (GAO) Cost Estimating and Assessment Guide, GAO-20-195G, published in March 2020, supersedes the 2009 edition and lays out a 12-step process. The Office of Cost Assessment and Program Evaluation (CAPE) publishes the DoD Cost Estimating Guide, Version 3.0, dated September 2025, which reflects policy and practice as of March 2025. The guide's foreword counts about 1,500 government cost analysts supporting an annual budget of more than $700 billion.

The 12 steps and four characteristics

The GAO guide treats a reliable estimate as one that is comprehensive, well documented, accurate, and credible. Comprehensive means the estimate defines the whole program, reflects the current schedule and technical baseline, and neither omits nor double-counts cost elements. Well documented means an analyst unfamiliar with the program could trace each element to its source and replicate it. Accurate means each element uses the best method for the data, is adjusted for inflation, and is updated as the program changes. Credible means the estimate states its limits, tests sensitivity, quantifies risk and uncertainty, cross-checks major elements, and is compared with an independent estimate made outside the acquiring organization.

The 12 steps are: define the estimate's purpose; develop the estimating plan; define the program; determine the estimating structure; identify ground rules and assumptions; obtain the data; develop the point estimate; conduct sensitivity testing; conduct a risk and uncertainty analysis; document the estimate; present it to management for approval; and update it to reflect actual costs and changes. GAO pairs the steps with 18 best practices. Step 2 has none, because a plan is not an attribute of the finished estimate. The steps need not run in order, and GAO advises documenting throughout.

How GAO applies the test

Auditors rate the estimate against each best practice, then summarize the ratings by characteristic. An estimate that fully reflects all four is considered reliable. A case study in the guide, drawn from GAO's June 2017 report GAO-17-575, applies the test to the Navy's Ford-class aircraft carriers. Construction costs for the lead ship, Gerald R. Ford (CVN 78), rose from $10.5 billion to $12.9 billion, nearly 23 percent. GAO found that the $11.4 billion budgeted for the follow-on ship, CVN 79, was likely insufficient. The estimate was substantially comprehensive. It was only partially well documented, partially accurate because of an optimistic assessment of labor hours, and partially credible because it did not sufficiently account for program risks. GAO concluded it was not a reliable basis for budgets, requirement trade-offs, or gauging shipbuilder progress.

CAPE's role in law and policy

10 U.S.C. 3221, in the 2024 edition of the U.S. Code, makes the CAPE director responsible for accurate cost information on DoD acquisition programs. The director prescribes cost estimating policy and reviews all cost estimates for major defense acquisition programs (MDAPs). The director also conducts or approves independent cost estimates (ICEs) before Milestone B approval, before low-rate initial or full-rate production, before a certification after unit cost growth reaches the critical threshold, and before a military construction contract above $500,000,000. Public Law 118-159 of December 23, 2024 struck the Milestone A reference from that list.

DoD Instruction (DoDI) 5000.73, effective October 24, 2024, predates that change and still lists Milestone A. The instruction defines the ICE as a full life-cycle estimate of development, production, deployment, operations and support, disposal, and military construction, without regard to funding source or management control. A milestone decision authority may not approve entering a milestone phase of an MDAP unless it has considered an ICE conducted or approved by the CAPE director. The CAPE guide lists the other types: the program office estimate, the DoD Component cost estimate, and the Component cost position that reconciles the two.

Life-cycle cost categories

CAPE's Operating and Support Cost-Estimating Guide, dated January 2025, states the DoD convention: life-cycle cost is the sum of four categories. They are research, development, test and evaluation (RDT&E); investment, meaning procurement plus military construction; operating and support (O&S); and disposal. For a major capability acquisition, RDT&E falls in the three phases before production, investment in production and deployment, O&S in operations and support, and disposal at the end. For many types of programs, O&S is the largest of the four.

The guide divides O&S into five top-level elements: unit-level manpower, unit operations, maintenance, sustaining support, and continuing system improvements. Indirect support costs, such as base operations support, are left out of O&S estimates made for acquisition decisions. Those estimates are checked again in sustainment. In an April 2026 report, GAO found that DoD identified O&S cost growth it classes as "critical" in 14 of 36 sustainment reviews for fiscal years 2023 and 2024. That label means at least 25 percent above the most recent ICE, or at least 50 percent above the original baseline estimate.

What to check

• Which editions does the estimate follow: GAO-20-195G of March 2020, and the DoD Cost Estimating Guide Version 3.0 of September 2025?
• Could an analyst unfamiliar with the program replicate each element from the documentation, as GAO's well-documented test requires?
• Does the estimate include a sensitivity analysis and a risk and uncertainty analysis that states a confidence level?
• Was the point estimate compared with an independent cost estimate made outside the acquiring organization?
• Does the life-cycle total include all four categories, disposal included, without regard to funding source?
• For an MDAP, was the ICE the milestone decision authority considered conducted or approved by CAPE?

Azcuna's LEDGER turns a trade study into a Decision Package in which every number cites its source and AI agents draft but never decide. It was proposed to the Army under SBIR Phase I topic ARM26BX06-NV012, and the proposal was submitted on October 21, 2026.

Sources

1. 1GAO-20-195G, Cost Estimating and Assessment Guide: Best Practices for Developing and Managing Program Costs · U.S. Government Accountability Office, March 2020
2. 2DoD Cost Estimating Guide, Version 3.0 · Office of the Secretary of Defense, Cost Assessment and Program Evaluation, September 2025
3. 3United States Code, 2024 Edition, Title 10, Section 3221 - Director of Cost Assessment and Program Evaluation · U.S. Government Publishing Office
4. 4DoD Instruction 5000.73, Cost Analysis Guidance and Procedures, October 24, 2024 · Office of the Director, Cost Assessment and Program Evaluation
5. 5Operating and Support Cost-Estimating Guide · Office of the Secretary of Defense, Cost Assessment and Program Evaluation, January 2025
6. 6GAO-26-108140, Weapon System Sustainment: DOD Identified Critical Cost Growth, and the Army Should Take Action to Yield Cost Savings · U.S. Government Accountability Office, April 23, 2026

Where this shows up in our work

• LEDGER, auditable Army acquisition decisions turns a trade study into a package where every number cites its source and AI drafts but never decides.
• ELLMENT, the offline LLM for E-2D crews answers a crew's question from the aircraft's own manuals and cites the page, with no network.
• TRACE, the offline S1000D and IETM viewer opens S1000D, MIL-STD-3001 and PDF technical manuals on a tablet with every marking intact.