DURIP FY2027 Carries $28 Million and Closes November 6 — Six Million Less Than Last Year, Ten Days Before NSF MRI

The Department of War released its FY2027 Defense University Research Instrumentation Program solicitation on July 29, 2026. Full proposals are due November 6, 2026 at 5:00 PM Eastern. Pre-proposal inquiries to the Service program managers close earlier — October 16, 2026 — which means that for most institutions reading this, the conversation phase is already over and the writing phase is all that remains.

The numbers that matter:

The single most consequential detail is not in the table. It is the delta: $28 million is roughly $6 million below FY2026, a decline of about 18 percent in a program that was already one of the smallest competitive instrumentation lines in the federal portfolio. DURIP is not being cut dramatically. It is being quietly compressed at the exact moment demand for instrumentation money is spiking.

What DURIP actually buys — and the three things it refuses to buy

DURIP is an equipment program in the strictest sense available in federal research funding. It funds "major, state-of-the-art equipment and instrumentation" in support of defense-relevant research. It does not fund:

• Direct research costs. No consumables tied to a research plan, no sample acquisition, no travel to conduct the work.
• Personnel. No PI salary, no student salaries, no postdoc support, no technician lines.
• General-purpose computing. A rack of standard servers or a general campus cluster expansion is outside scope.

That third exclusion deserves attention in October 2026, because it cuts directly against the direction the rest of the federal research enterprise is moving. The month has been dominated by compute: $2.4 billion in industry compute pledges to the Genesis Mission and the regional consortia forming around them, NSF's state and regional AI infrastructure hubs, DOE's quantum and autonomous-laboratory lines. A PI who reads that environment and concludes that DURIP is the place to ask for GPUs has misread the solicitation. DURIP wants a scientific instrument — a spectrometer, a chamber, a beam line component, a characterization system, a specialized sensor array. Computing enters a competitive DURIP proposal only as the control and acquisition layer attached to a physical instrument, not as the thing being purchased.

The practical translation: if your equipment need is compute, DURIP is the wrong door and there are now several right ones. If your equipment need is a physical instrument, DURIP is one of the very few doors left that will pay for the instrument alone, with no requirement that you simultaneously win a research grant to justify it.

The three-service routing problem

DURIP is not one competition. It is three, running under a shared framework, and the proposal goes to exactly one of them:

• Army Research Office (ARO) — engineering sciences, physical sciences, life sciences, information sciences
• Office of Naval Research (ONR) — warfighter performance, cyber, AI/ML, materials, ocean and atmospheric sciences
• Air Force Office of Scientific Research (AFOSR) — engineering and physical sciences, bio-cognitive sciences

Routing is a scoring decision disguised as an administrative one. The Services evaluate against their own portfolio interests, and an instrument that is merely interesting to one Service may be central to another's roadmap. A materials characterization system framed for ARO's engineering sciences portfolio and the same system framed for ONR's materials portfolio are not equally competitive — they are competing against entirely different stacks of proposals, with different program officers and different unfunded priority lists.

This is why the October 16 inquiry deadline was the real deadline for a large share of applicants. The Services ask PIs to contact the relevant program officer before drafting, precisely so that routing gets settled before a month of writing is committed. If you missed it, you are not disqualified — the inquiry window is for clarifying topic alignment, not a gating requirement — but you are now making the routing call on your own inference from published portfolio descriptions rather than on a program officer's guidance. Read the Service BAAs' research-area taxonomies closely and route to the office whose language your instrument's application already speaks.

The November collision: DURIP and NSF MRI, ten days apart

The strategically important fact about DURIP FY2027 is its calendar position. NSF's Major Research Instrumentation program reopened October 15, 2026 after skipping an entire year, and closes November 16, 2026 with roughly $75 million against roughly 100 awards. DURIP closes November 6. The two windows overlap almost completely, and most research offices are staffing both with the same people.

They are not interchangeable programs, and the differences dictate which instrument goes where:

Four implications follow.

First, DURIP has no stated per-institution proposal cap while MRI caps at four. For a large research university with a long instrumentation queue, DURIP is the release valve. The MRI four-proposal cap forces brutal internal triage; DURIP does not impose the same arithmetic. Instruments that lose an internal MRI competition are not dead — they are DURIP candidates, provided a defense application is genuine rather than retrofitted.

Second, the small end of DURIP is underused. The floor is $50,000. Most institutional attention goes to seven-figure asks, but a $28 million pool funding awards as small as $50,000 can support a large number of modest instruments. A $120,000 request for a single well-justified piece of characterization equipment is a real and often under-competed position, and it is far below the threshold where MRI is worth the proposal effort.

Third, no cost share means no institutional negotiation. DURIP requires none. MRI's waiver is in place for FY2027 but has been signaled as the final year. For an institution whose matching capacity is already committed, DURIP is the cleaner path in FY2027 — and it will be the only clean path in FY2028 if the MRI waiver lapses as expected.

Fourth, the education requirement is scored, not decorative. Competitive DURIP proposals demonstrate impact on both defense-relevant research and student education. The common failure is a proposal that justifies the instrument entirely through the PI's research program, with a single sentence about graduate student access. Name the courses. Name the number of students per year who will run the instrument themselves rather than hand samples to a staff operator. Name the undergraduate research pathway it opens.

The contraction context

DURIP's $6 million reduction is not an isolated budget decision. Federal grant notice volume has been through a structural reset — posted opportunities fell by more than half from 2025 into 2026 — and the instrumentation lines have been hit in a particular way. MRI skipped an entire fiscal year. NSF's governing MRI solicitation is archived with no published replacement. DURIP is down 18 percent. Meanwhile the new money in the system is flowing toward AI-enabled instrumentation and autonomous laboratories, where NSF and DOE together announced more than $100 million at the October science summit, and toward compute access allocated through consortia rather than competed through instrumentation grants.

The net effect for a PI who needs a physical instrument is that the traditional equipment pathways have gotten narrower while the new pathways reward a specific kind of instrument: one that is automated, data-generating, and plausibly a node in an AI-driven discovery loop. That is not a reason to rewrite a vibration-isolation table as a machine-learning platform. It is a reason, where it is honest, to describe what the instrument's data stream enables downstream — because the reviewers reading DURIP proposals this November are reading them inside an agency environment saturated with exactly that framing.

What to do in the next three weeks

Confirm your Service routing from published portfolio language, then stop second-guessing it. With the October 16 inquiry window closed, the cost of changing routing mid-draft is higher than the expected gain. Pick based on which office's research-area description your instrument's application matches most literally.

Get a quote with a validity date past the award cycle. Equipment proposals die on stale pricing. Vendor quotes that expire in December create an awkward conversation in a March negotiation.

Write the shared-use plan even though DURIP does not require one the way MRI does. An instrument that serves three PIs and two departments is a better defense investment than an instrument that serves one lab, and reviewers across both programs read it the same way.

Split your queue deliberately rather than sequentially. The instinct is to send the top instrument to MRI and the leftovers to DURIP. The better logic routes by fit: defense-relevant, single-lab-intensive, and under $500,000 goes to DURIP; broadly shared, campus-scale, and education-forward goes to MRI. An instrument with a real Service champion is more competitive at $28 million against a defense-screened field than at $75 million against a national open field.

Treat the November 6 date as November 5. A 5:00 PM Eastern deadline in a Service submission system, during a month when your research office is also processing MRI proposals due ten days later, is not a date to test.

The honest read on DURIP FY2027 is that it is a shrinking program that remains one of the best-value propositions in federal research funding: no cost share, no personnel burden to justify, a $50,000 floor, and no per-institution cap in a year when its only real alternative caps at four. The money is smaller than last year. The arithmetic is still favorable.