How a Pentagon Program Moves From Requirement to Fielding

A Pentagon acquisition program moves from requirement to fielding through a gated process run jointly by the military services and the Office of the Secretary of Defense: a documented requirement, a Milestone B decision that starts engineering and manufacturing development, a Milestone C decision that starts low-rate production, and a full-rate production decision that follows operational testing. The Government Accountability Office's annual assessments repeatedly find that programs spend longer in development than planned — weapons programs GAO reviewed in its 2023 report delivered capabilities roughly two years late on average, per GAO-23-106056.

Advanced Primitive publishes open-source analysis, not advice, and every fact below is attributed to its document type.

Where does a program actually begin?

It begins with a requirement, not a sketch. A combatant command or service identifies a capability gap, and the staff writes it into the Joint Capabilities Integration and Development System — the formal process, abbreviated JCIDS, that validates whether the need is real and whether an existing system already fills it. Most proposed requirements die here, in analysis, before any money moves.

A validated requirement becomes a program of record only when a Milestone Decision Authority approves an acquisition approach. That authority sits with the service acquisition executive for most programs and with the Office of the Secretary of Defense for the largest ones — those designated Major Defense Acquisition Programs, or MDAPs, a label that attaches mandatory oversight, reporting, and exit criteria.

What are the milestones, in plain terms?

The milestone system is a chain of gated decisions, each with entry criteria written down before the gate is approached. The canonical sequence for a major program:

1. Milestone A — approves the technology-maturation and risk-reduction phase; the program may spend money on prototypes and trades, not on production.
2. Milestone B — approves engineering and manufacturing development; this is the formal start of system development, with a cost and schedule baseline that Congress is notified about.
3. Milestone C — approves low-rate initial production (LRIP), enough units to complete operational testing and establish the production line.
4. Full-rate production — approved after operational test reports and a production-readiness review; the program is now buying at scale.

The gates exist because each phase buys information the next one needs. Skipping a gate — usually under schedule pressure — is a documented driver of the cost growth that GAO tracks: entering production before testing is complete means finding defects on the production line, where each fix is more expensive.

Why does development take so long?

Because the schedule is set by concurrent learning, not by assembly. During engineering and manufacturing development, the design, the factory process, and the support system — spares, trainers, technical orders — all have to converge before production can start. A fighter that flies is not a fielded system; a fielded system has maintainers trained, software baselines stable, and a depot planned.

Software now sets the critical path more often than airframes do. GAO's 2023 assessment found that major programs were averaging years between Milestone B and fielding, with concurrency between testing and production the recurring structural risk — the practice of building production units while test results are still arriving, which delivers hardware earlier at the price of retrofits.

Who approves what, and when does Congress see it?

Approval authority scales with cost. Programs expected to cost more than statutory thresholds in research or procurement dollars — hundreds of millions, with the exact threshold adjusted for inflation and stated in 10 U.S.C. § 4031's predecessor framework — are designated MDAPs and face the full oversight stack: service review, OSD review, and congressional notification of the baseline at Milestone B.

Congress sees money before it sees hardware. A program cannot obligate funds until its budget line survives the annual appropriations cycle, and a continuing resolution — funding at prior-year levels in the absence of an appropriations bill — blocks new program starts, a documented source of schedule slips in GAO and Congressional Research Service reporting.

What usually goes wrong?

Three patterns recur across decades of GAO weapons assessments. Requirements growth: the document that survived Milestone B keeps acquiring features, and each one moves the baseline. Schedule optimism: the promised Milestone B-to-C interval assumes no major redesign, which testing routinely disproves. Production concurrency: LRIP quantities arrive before operational testing finishes, so early aircraft need changes at the depot.

One dry aside from the record: GAO has reported on the same structural findings — optimistic baselines, concurrent testing and production — every year since the 1980s; the acquisition reporting framework itself has been reorganized at least three times since then.

What remains unknown for any specific program is always the same: whether the current baseline holds. The document trail states the promised dates; the milestone history states what happened to the last set. Reading them side by side is the whole craft of tracking a program.