NSF's Instrumentation Program Reopens October 15 After Skipping an Entire Year — Under a Solicitation NSF Has Already Archived

On October 15, 2026, the National Science Foundation's Major Research Instrumentation Program begins accepting new proposals for the first time since November 2024. The window runs through November 16, 2026. It is the program's FY2027 competition, and it is the single most consequential shared-equipment competition in American academic science — roughly $75 million against roughly 100 awards, at a ceiling of $4 million per proposal.

Three things make this particular opening unusual, and all three change how an institution should approach it.

The window exists because the previous one was cancelled outright. The document that governs it has been formally archived by NSF. And it is the final competition covered by a statutory cost-sharing waiver that expires with FY2027 — which means the next MRI competition after this one may well arrive with a 30 percent match attached.

A Program That Skipped a Year

MRI normally runs annually on a fall window. The FY2025 competition accepted proposals October 15 – November 15, 2024.

Then, on July 1, 2025, NSF published a program update stating it would not accept new proposals during the already-announced FY2026 window of October 15 – November 14, 2025. The language was unambiguous: "Any proposals submitted during this period will be returned without review."

The stated reason was a backlog, not a cut. NSF said it held many meritorious proposals from the FY2025 cohort and would, subject to available funds, use FY2026 money to make additional awards from that existing pile rather than run a fresh competition.

That is a defensible use of money in a constrained year. It is also an invisible one-year suspension of the primary federal mechanism by which universities buy cryo-EMs, mass spectrometers, NMR systems, high-performance computing clusters, and research vessels. Instruments do not stop aging because a window closed. Equipment that was marginal in late 2024 is now two years further into obsolescence, and the demand that would have been spread across two competitions is about to arrive in one.

Expect this competition to be the most oversubscribed MRI round in years. Not because NSF changed anything about merit review, but because the queue has had 23 months to fill.

The Archived-Solicitation Problem

Here is the part that will confuse sponsored-programs offices, and it is worth being precise about.

The governing solicitation is NSF 23-519. Its own submission-window table lists October 15 – November 16, 2026, and carries the explicit note that the October 15 – November 14, 2025 window was cancelled. So the date is in the document.

But NSF 23-519 is labeled, at the top of its own page, "This solicitation is archived." And the MRI program page states that NSF "anticipates issuing an updated MRI solicitation for an opportunity to submit proposals in FY 2027."

That update was expected "later in FY 2026." Fiscal year 2026 ended on September 30, 2026. As of this writing — four days before the window opens — no replacement solicitation has been published.

Universities noticed. The University of Colorado Boulder's internal MRI call carries an NSF deadline of "TBD" and the caveat that the call is subject to change based on updated NSF guidance "expected later in 2026." Boise State's limited-submission notice names the October 15 – November 16, 2026 NSF window while stating that details "may be subject to change with publication of new solicitation."

The practical reading: plan against NSF 23-519's rules, because they are the only published rules, and build in the assumption that a new solicitation could drop inside the window with changed requirements. A solicitation released on, say, October 20 with a modified project-description structure would leave four weeks to rebuild documents that institutions have been drafting since August. Keep the science narrative modular and the compliance sections thin enough to swap.

The Real Competition Already Happened

MRI is a limited-submission program, and this is where most would-be applicants lose before NSF ever sees a page.

Under NSF 23-519, each performing organization may submit:

• Track 1 — NSF request greater than $100,000 and less than $1,400,000: maximum two proposals
• Track 2 — NSF request from $1,400,000 up to and including $4,000,000: maximum one proposal
• Track 3 — helium-conservation equipment, $100,000 to $4,000,000: maximum one proposal, which does not count against Tracks 1 and 2

Four slots, total, for an entire university. Track 3 exists because of section 10373 of the CHIPS and Science Act and specifically targets equipment that conserves or reduces helium consumption.

Because the cap is institutional, every research university runs an internal competition months ahead of NSF's window. Boise State's white papers were due August 13, 2026, with internal selections announced September 2. CU Boulder's internal deadline was August 17, 2026.

If you are a PI reading this on October 11 without an internal slot, you are not applying this cycle. That is the honest answer, and it is better to hear it now than after assembling a budget. The useful move is to find out who holds your institution's four slots, ask whether any assignment is soft, and — more durably — get your instrument in front of your research office before next summer's internal call.

There is one real exception worth checking: the submission limit is per performing organization. A collaborative proposal where your instrument lives at a partner institution with an unused slot is a legitimate structure, not a loophole. Multi-institutional access plans also score well on their own merits, since MRI explicitly rewards instruments that serve users beyond a single campus, including users from non-Ph.D.-granting institutions.

FY2027 Is the Last Year Without a Match

This is the strategic fact that almost nobody is pricing in.

Cost sharing in MRI is not an agency preference. It is statute. Section 7036(c) of the America COMPETES Act of 2007 required cost sharing on MRI awards, and under the pre-waiver rules — NSF 15-504 is a clean example — the requirement was precisely 30 percent of total project cost for Ph.D.-granting institutions and non-degree-granting institutions. Non-Ph.D.-granting institutions of higher education were exempt and were prohibited from providing it.

Section 10320 of the CHIPS and Science Act of 2022 waived that requirement for a period of five years, beginning with the FY2023 MRI competition.

Count the competitions: FY2023, FY2024, FY2025, FY2026, FY2027. The window opening October 15 is the FY2027 competition — the fifth and final year of the waiver.

What that means in dollars: a $4 million instrument today is a $4 million NSF request with no institutional match. Under the restored statutory rule, the same instrument becomes a $5.71 million project in which NSF pays exactly 70 percent — $4 million — and a Ph.D.-granting institution puts up exactly $1.71 million of its own money. Nothing about the instrument changes. The institution's share goes from zero to $1.71 million.

Congress could extend the waiver. NSF's forthcoming solicitation could address it. Neither has happened yet. Until one does, any Ph.D.-granting institution with a large instrument need should treat this window as materially cheaper than the next one and weight its internal slot allocation accordingly. A Track 2 proposal deferred to FY2028 is a proposal that may need a seven-figure match that does not currently exist in anyone's budget.

One related compliance trap in the meantime: because the requirement is waived, voluntary committed cost sharing is prohibited by National Science Board policy. Do not volunteer institutional funds to look stronger. Since voluntary cost sharing is not permitted, the maximum total project cost equals the maximum NSF request — $4 million, full stop.

Who Is Eligible, and Who the Program Is Actually Courting

Eligible submitters are U.S. institutions of higher education — both Ph.D.-granting and non-Ph.D.-granting — not-for-profit scientific or engineering research organizations, and consortia of eligible organizations.

There is a genuine carve-out at the low end. Track 1 requests below $100,000 are accepted only from eligible organizations seeking instrumentation for mathematics or social, behavioral and economic sciences, or from non-Ph.D.-granting institutions across any NSF-supported discipline. For a primarily undergraduate institution, that is a real and under-used lane: a sub-$100,000 request that a research-intensive university simply cannot make.

NSF 23-519 also states plainly that MRI "especially seeks broad representation of groups, institutions, and geographic regions that are underrepresented in STEM disciplines," and encourages proposals from under-resourced and emerging research institutions, from EPSCoR jurisdictions, and from early-career PIs. The solicitation separately encourages proposals advancing U.S. leadership in microelectronics research and training, under section 10318(a)(6)(F) of CHIPS and Science.

Read against the 23-519 rules, the sharpest positioning available to a smaller institution is a Track 1 acquisition proposal with a concrete external-user access plan, filed from an institution that faces no internal slot competition at all.

What To Do Between Now and November 16

1. Confirm which solicitation governs before you build documents. Check the MRI program page for a replacement solicitation daily through the window. If one appears, re-verify the project-description requirements and the PAPPG version it invokes.
2. Confirm you hold an institutional slot. Four per campus. Ask your sponsored-programs office in writing.
3. Classify your organization correctly. 23-519 requires a signed statement from the sponsored projects office classifying the performing organization as non-Ph.D.-granting, Ph.D.-granting, or non-degree-granting. Subawardees need it too.
4. Strip any voluntary cost sharing. It is prohibited, and including it is a returned-without-review risk.
5. Price the FY2028 alternative honestly. If the waiver lapses, deferring a $4 million Track 2 request costs your institution roughly $1.71 million. That calculation belongs in your internal slot argument, not in the proposal.

MRI has been running since the 1990s and has survived budget cycles far worse than this one. But a program that cancels a window, archives its own solicitation, misses the publication date for the replacement, and enters the final year of a cost-share waiver all at once is a program worth treating as a closing door rather than an annual fixture.

For context on the budget environment MRI is competing inside, see our analysis of NSF's FY2026 close-out and the $1.7 billion carryover and of the $1 billion central account behind Grand Research Challenges.