NSF Analysis Guide
Updated 20 official sources
Quick Answer: NSF Analysis eligibility now runs through NSF 26-520 (opens in a new tab), posted August 17, 2026, which accepts proposals anytime.
Full answer and how we checked it
If you're a PI at a US college, university, or non-profit research organization, the old PD 20-1281 September window is gone. Grantsights analysis of NSF award records puts the FY2026 median Analysis research award at $200,000, most often over 3 years, so the real questions are fit, route, and a budget that matches what the program now funds.
Who this is for: Analysis PIs, PDE researchers, mathematical physics and operator-algebra groups, sponsored-research offices, and workshop organizers deciding whether NSF Analysis is the right home and how big a request looks normal in 2026.
Data note: This guide uses the official NSF Analysis page, the NSF 26-520 solicitation, Grants.gov listing 325500 (opens in a new tab), the NSF PAPPG page and its 2025 and 2026 supplements, NSF's priorities FAQ, and the NSF Award Search API (opens in a new tab). Program status, policy pages, and award data were checked on October 18, 2026. Award records were pulled on October 18, 2026 in award-date windows small enough that each request returned every matching record, because paging through larger API result sets can repeat some records and skip others.
In this guide
- 7 NSF Analysis Eligibility Stop Checks
- 6 Source Facts Applicants Should Know
- What Changed From PD 20-1281 to NSF 26-520
- NSF-Wide Rules That Changed in 2025 and 2026
- What NSF Doesn't Tell You
- What Recent Analysis Awards Look Like
- CAREER, FRG, and Conference Routes
- 6 Evidence Assets That Make This Proposal Stronger
- 5 Route Choices Before Drafting
- 6 Bad-Fit Signals
- Before You Submit: 5 Final Checks
- Sources
Quick Facts
- Agency
- NSF
- Last deadline
- Sep 30, 2026 (passed)
- Status
- Closed
Assistance Listing 47.049: all awards
Source: USASpending.gov FY2024These figures cover all U.S. National Science Foundation awards under Assistance Listing 47.049, of every award type. They are not figures for the award this guide describes, and one award of that kind can be far smaller.
- Total Awards (FY2024, whole listing)
- 5,546
- Avg Award (whole listing)
- $1,395,751
- Median Award (whole listing)
- $375,000
- Total Funding (FY2024, whole listing)
- $1.6B
- YoY Funding Trend
- -7.8%
Data last verified: May 19, 2026
According to NSF Analysis (opens in a new tab), the program supports complex, harmonic, and real analysis; dynamical systems and ergodic theory; functional analysis; mathematical physics; operator theory and operator algebras; partial differential equations; and calculus of variations. NSF says that list isn't closed, and proposals on other relevant analysis topics will also be considered.
What changed in 2026 is the submission route. Analysis used to run on its own program description, PD 20-1281, with a September 16 to 30 window. On August 17, 2026, NSF posted NSF 26-520, a single solicitation that "consolidates most mathematical sciences funding opportunities" in MPS, Analysis included, with proposals accepted anytime. The PD 20-1281 listing on Grants.gov (opens in a new tab) closed September 30, 2026.
The decision question is still route fit: does the mathematical contribution belong in Analysis, or would Geometric Analysis, Probability, Applied Mathematics, Computational Mathematics, Statistics, Topology, CAREER, or conference support be a better home? Search the Grantsights grant database for the live program record and nearby math programs.
7 NSF Analysis Eligibility Stop Checks
Seven checks settle program fit before anyone reads the theorems. NSF 26-520 limits submitters to 3 groups: accredited US two- and four-year institutions of higher education acting for their faculty, US non-profit non-academic organizations tied to research or education (museums, observatories, research labs, professional societies), and federally recognized tribes. It places no restriction on who may serve as PI.
| Analysis check | Signal the program fits | Red flag for this program |
|---|---|---|
| Analysis core | Analysis is the main contribution | Analysis is only a tool |
| Subfield | Fits one of NSF's listed areas | Better in another math program |
| Solicitation | Proposal is prepared under NSF 26-520 | Draft still follows PD 20-1281 instructions |
| Conference | Meeting support is a separate proposal | Workshop ask is hidden in research |
| Route | Analysis is the best home | Geometric Analysis or Probability fits better |
| Budget | Request sits near the $200,000 to $300,000 band | Single-PI ask well above $300,000 with no reason |
| Broader impacts | Training or math-community value is specific and open to all | Generic outreach only |
A common mistake is writing for every math panel at once. Analysis proposals weaken when the introduction tries to sound like geometry, probability, and applied math in the same paragraph. Pick the home, then write for it. The strongest drafts name the subfield and the main theorem on page 1.
6 Source Facts Applicants Should Know
Grantsights checked NSF 26-520 line by line against the Analysis program page and the retired Grants.gov listing (opens in a new tab). These 6 facts drive most planning decisions.
First, the scope. According to NSF, Analysis covers complex, harmonic, and real analysis; dynamical systems and ergodic theory; functional analysis; mathematical physics; operator theory and operator algebras; partial differential equations; and calculus of variations.
Second, the scale. NSF 26-520 lists an estimated 400 to 600 awards and $200,000,000 across all 14 MPS math programs it covers, not just Analysis. That figure is shared with Algebra and Number Theory, Applied Mathematics, Combinatorics, Computational Mathematics, Foundations, Geometric Analysis, Mathematical Biology, Probability, Statistics, Topology, Focused Research Groups, Research Training Groups, and the Infrastructure program.
Third, the award types. According to NSF 26-520, Analysis proposals can become standard or continuing grants; the solicitation prohibits voluntary committed cost sharing and lists "Not Applicable" for indirect cost limitations.
Fourth, length. According to NSF, the agency now encourages proposals requesting up to 5 years of support where longer awards would help ambitious work, and it says budgets and durations must be commensurate with scope.
Fifth, submission systems. You can submit through Research.gov or Grants.gov (opens in a new tab), but collaborative proposals sent as separate submissions from multiple organizations must go through Research.gov (opens in a new tab).
Sixth, the money source. According to USAspending.gov (opens in a new tab) data checked October 18, 2026, assistance listing 47.049 had about $1.60B in FY2024 obligations, with 5,656 grant awards showing FY2024 activity. That figure covers all of NSF's Mathematical and Physical Sciences directorate, so it's context, not an Analysis budget.
| Fact | Source-backed read |
|---|---|
| Current solicitation | NSF 26-520 (MPS MathSci), posted August 17, 2026 |
| Deadline | Proposals accepted anytime, 5 p.m. submitter's local time |
| Older program description | PD 20-1281; Grants.gov listing 325500 (opens in a new tab) closed September 30, 2026 |
| Eligible submitters | US IHEs, US non-profit research organizations, federally recognized tribes |
| Award types | Standard or continuing grant; up to 5 years encouraged where justified |
| Cost sharing | Voluntary committed cost sharing prohibited |
| Organization | Directorate for Mathematical and Physical Sciences |
| Contact | Analysis@nsf.gov (program directors Marian Bocea and Wing Suet Li) |
What Changed From PD 20-1281 to NSF 26-520
At least 7 rules moved when Analysis left PD 20-1281. The old Grants.gov listing (opens in a new tab) told conference organizers to apply through a separate "Conferences and Workshops in the Mathematical Sciences" solicitation 8 months before the event. NSF 26-520 replaces that with a 6-month lead time for requests up to $50,000 and 9 months above $50,000.
The other changes are in the solicitation text itself:
- No window. Proposals are accepted anytime instead of during a September 16 to 30 window.
- One solicitation for most math programs. Focused Research Groups and Research Training Groups now sit in the same document, so interdisciplinary proposals that span more than one math program are explicitly welcome.
- Formal mathematics and AI. NSF encourages work that puts mathematical knowledge in machine-readable form for sharing and verification, and proposals that use AI to speed mathematical discovery, as long as the use is "scientifically justified rather than incidental."
- Longer awards. The solicitation encourages up to 5 years of support where it fits the science.
- Indirect-cost preference. Citing Executive Order 14332, NSF says that "all else being equal after merit review" it may give preference to institutions with lower indirect cost rates.
- Termination for convenience. Awards under NSF 26-520 may include termination-for-convenience provisions.
- Review mechanics in writing. The solicitation states that at least 2 primary reviewers are assigned to each research proposal and that triage may be used.
The practical takeaway: don't reuse a PD 20-1281 cover page, budget justification, or conference timeline without rechecking each one against NSF 26-520.
NSF-Wide Rules That Changed in 2025 and 2026
Four NSF-wide policies affect an Analysis proposal written in late 2026, from a vacated 15% indirect cost cap to a new data-plan tool, and none of them appear on the Analysis program page.
The PAPPG in effect is still NSF 24-1. According to the NSF PAPPG page (opens in a new tab), NSF 24-1 applies to all proposals submitted or due on or after May 20, 2024, and there's no NSF 25-1 in its version history. Supplement 1 (NSF 26-200) (opens in a new tab), effective December 8, 2025, says NSF deferred release of the planned NSF 26-1 PAPPG after Executive Order 14332 and will issue interim changes as policy notices. It also says a full proposal must be reviewed by at least 2 reviewers, and that NSF staff can conduct those reviews internally, and that a proposal can be returned without review if it lacks enough content for merit review. NSF also posted a draft "Guidance on Financial Assistance" meant to replace the PAPPG, with comments that closed August 24, 2026; until NSF publishes a final version, NSF 24-1 plus supplements governs.
Data and publication rules tightened. Supplement 2 (NSF 26-202) (opens in a new tab), effective January 22, 2026, requires the Data Management and Sharing Plan to be built in a Research.gov tool (released April 27, 2026) instead of a two-page PDF upload. Recipients must share the data supporting NSF-funded publications at the time of publication, and author accepted manuscripts must be deposited in NSF's Public Access Repository at or before publication. For a pure analysis project, a short plan that says no datasets will be produced is still valid, but it needs a clear justification and should say where preprints, code, or formalized proofs will live.
The 15% indirect cost cap isn't in effect. According to NSF's priorities FAQ (opens in a new tab), a federal court vacated Policy Notice NSF 25-034, which set a standard 15% indirect cost rate for universities, and NSF has issued awards at negotiated rates since May 19, 2025. The court record summarized by the Civil Rights Litigation Clearinghouse (opens in a new tab) shows a June 20, 2025 final judgment in Massachusetts federal court and a government appeal that was voluntarily dismissed on September 30, 2025. Two caveats remain: NSF says new awards carry a term applying NSF 25-034 if it's later allowed to, and NSF 26-520 adds the lower-rate preference described above. Budget at your negotiated rate, and don't cut it voluntarily without talking to your sponsored-research office.
Broader impacts are scored the same way but framed differently. NSF still reviews every proposal on Intellectual Merit and Broader Impacts. Its priorities FAQ tells investigators to prioritize the first 6 broader impacts goals defined in the America COMPETES Reauthorization Act of 2010, and says any activity aimed at goal 7, expanding participation of women and underrepresented groups, must be open to all Americans. Participation can still be expanded by non-protected traits such as institution type, geography, socioeconomic status, and career stage. For an Analysis PI, that points toward plans like an open summer school, an REU-style problem seminar, or graduate training tied to the research aims.
The org chart changed too. The old Division of Mathematical Sciences address (opens in a new tab) now redirects to an MPS Mathematical Sciences (opens in a new tab) page, and the Analysis program page lists only the MPS directorate. In NSF's award API, 287 of the 319 award records carrying the Analysis code now name the managing unit "Mathematical Science Section" instead of "Division Of Mathematical Sciences." Program contacts and the Analysis@nsf.gov address are unchanged.
What NSF Doesn't Tell You
Four practical gaps sit between the solicitation and a funded Analysis award, which had a $200,000 median in FY2026. First, the word analysis is too broad to decide fit. The proposal must show why the central contribution belongs in one of the Analysis areas NSF names, and the first page should say which one.
Second, "accepted anytime" removes the window but not the timing problem. Under the old September window, 42 of the 55 FY2026 research awards were dated June through August, and most started in July or August. NSF hasn't published a review calendar for anytime submissions, so ask the program directors before promising a start date to students or a postdoc.
Third, adjacent-route risk is real. A new harmonic-analysis estimate is a natural Analysis proposal. A project whose main output is a PDE solver for an engineering system usually belongs in Applied Mathematics, Computational Mathematics, or an engineering program. Co-review happens: in the award data, 32 of the 319 Analysis-coded awards were managed by another program, 29 of them by Applied Mathematics, Geometric Analysis, Topology, or Algebra and Number Theory, so a split proposal can still land, but only one program owns it.
Fourth, NSF tags awards in ways that show where the money is moving. In FY2026, NSF attached its artificial intelligence reference code to 16 of the 88 awards carrying the Analysis code and quantum information science codes to 18, up from 0 and 1 in FY2024. Those tags are NSF's own labels, not a scoring rule, but they line up with NSF 26-520's emphasis on AI and quantum. If an operator-algebra or harmonic-analysis project has a genuine link to quantum information or the mathematics of learning, state it plainly. Don't bolt on a connection that the aims don't support; the solicitation warns against incidental use.
What Recent Analysis Awards Look Like
Grantsights analysis of NSF Award Search API records, pulled October 18, 2026, covers 319 awards that list the Analysis program element (128100), including co-funded awards, dated October 1, 2023 through September 30, 2026. The pull searched by fund program name in award-date windows of 25 records or fewer, so no request relied on paging, then kept only records whose program element codes include 128100; two separate runs returned the same 319 award IDs. Awards were grouped by federal fiscal year of the award date and sorted into CAREER (NSF's CAREER reference), conferences (conference or workshop in the title, or NSF's conference reference code), Focused Research Groups, and standard research awards. Dollar figures are each award's estimated total, including any share paid by co-funding programs.
| Fiscal year | All awards | Research awards | Median research award | Research awards above $300,000 | 2-year research awards | CAREER | Conferences | Total value |
|---|---|---|---|---|---|---|---|---|
| FY2024 | 114 | 74 | $291,310 | 29 | 2 | 9 | 30 | $29.3M |
| FY2025 | 117 | 77 | $254,498 | 15 | 8 | 7 | 33 | $23.9M |
| FY2026 | 88 | 55 | $200,000 | 4 | 12 | 3 | 29 | $15.1M |
The finding that changes the budget: the typical Analysis research award fell by about a third in two years. The median dropped from $291,310 in FY2024 to $200,000 in FY2026, and the median per year fell from about $98,000 to about $80,000. Flat totals of $200,000 (13 awards) and $300,000 (9 awards) were the two most common FY2026 sizes. Only 4 of 55 FY2026 research awards topped $300,000: two were 1-year awards (Texas A&M, $378,967, and Emory, $339,651), and the other two carried EPSCoR co-funding (University of Alabama, $327,888, and University of Delaware, $327,908). The total value of Analysis-coded awards fell 48%, from $29.3M to $15.1M.
Durations shortened too. Twelve FY2026 research awards ran 2 years, compared with 2 in FY2024, and continuing grants nearly vanished (13 research awards in FY2024, 2 in FY2026). That matters if you were planning a 3-year continuing grant with student support in every year.
Recipients reach well beyond the largest research universities. The 88 FY2026 awards went to 55 different institutions in 31 states and the District of Columbia. Rutgers University New Brunswick and the University of Maryland, College Park had 5 each, and UCLA had 4. Specifically, Saint Louis University won $200,000 over 3 years for phase retrieval and nonlinear frame theory (2556024 (opens in a new tab)), the University of Louisiana at Lafayette won $234,698 for work on C*-algebras (2555075 (opens in a new tab)), and Ball State University won $125,349 for mappings in non-smooth geometry (2553261 (opens in a new tab)).
EPSCoR co-funding shows up often. Thirteen of the 55 FY2026 research awards carried NSF's EPSCoR reference, at a median of $234,698, above the overall median. Recipients included the University of Hawaii, Auburn University, Tulane University, Kansas State University, the University of Oklahoma, and Clemson University. If your campus is in an EPSCoR jurisdiction, read the NSF EPSCoR guide before setting the budget.
The largest awards were CAREER and Focused Research Group projects. For example, Georgia Tech Research Corporation received a $450,000 CAREER award (2540992 (opens in a new tab)) for long-time behavior of dispersive waves, running 2026 to 2031, and the University of California-Berkeley share of a collaborative FRG on quasiperiodic operators (2555541 (opens in a new tab)) totals $566,385. All 3 FY2026 CAREER awards were 5-year awards of about $450,000.
The practical read: a single-investigator request well above $300,000 for 3 years now stands out against what the program funds. If you need more, change the budget justification to show why (several graduate students, a postdoc, or a collaborative structure), or consider a 2-year request at a smaller total. Compare your numbers with the NSF budget preparation guide before routing it.
CAREER, FRG, and Conference Routes
Three Analysis-adjacent routes run on different rules, and the dollar gaps are large: about $450,000 for a CAREER award, $566,385 for the Berkeley share of the one FY2026 FRG, and a median of $25,000 for an FY2026 conference.
CAREER is not in NSF 26-520. It runs on its own solicitation, NSF 22-586 (opens in a new tab), with a deadline on the fourth Wednesday in July; NSF's CAREER page lists the next one as July 28, 2027. For MPS, the CAREER award is expected to total at least $400,000 over 5 years including indirect costs, and each PI gets one CAREER submission per annual competition. An assistant professor with a 5-year research and education plan should usually apply to CAREER first and use NSF 26-520 for a standard grant in a different year. The NSF CAREER award guide covers eligibility.
Focused Research Groups now sit inside NSF 26-520. They fit collaborative groups attacking a specific major problem, not a loose collection of single-PI projects.
Conferences follow the Analysis page rules:
- Submit with the "conference" or "travel" proposal type in Research.gov and select Analysis as the program.
- File at least 6 months ahead for requests up to $50,000, and at least 9 months ahead above $50,000.
- Send meetings that span several math programs to the Mathematical Sciences Infrastructure program instead.
- Expect lighter review: external review isn't required for conference proposals up to $200,000.
The FY2026 conference awards were small and regional. The University of Richmond received $25,000 for the Southeastern Analysis Meeting, the University of Pittsburgh $12,000 for a harmonic analysis and PDE workshop, and the University of Kansas and Kansas State University $25,000 each as a collaborative award for the Prairie Analysis Seminar. The FY2026 conference median was $25,000, and only 2 quantum-computing meetings at UCLA, funded jointly with Applied Mathematics and Physics at the Information Frontier, went above $60,000.
6 Evidence Assets That Make This Proposal Stronger
Six documents, built before the project description, make an Analysis proposal easier to review.
Asset 1 is the fit paragraph. In 2 sentences, name the analysis area, the mathematical object, the theorem direction, and why Analysis rather than a neighboring program should review it.
Asset 2 is the subfield map. Place each aim in one of NSF's listed areas, such as harmonic analysis, dispersive PDE, or operator algebras, and note any aim that drifts into geometry or probability.
Asset 3 is the proof-path memo. Separate established lemmas, proposed methods, speculative aims, and fallback results, so reviewers can see what survives if the hardest step fails.
Asset 4 is the budget table checked against the award data above: total, years, students, and per-year cost compared with the FY2026 median of about $80,000 a year.
Asset 5 is the broader impacts plan, written to NSF's current framing: training, open seminars, courses, or research experiences, with participation open to all. The NSF broader impacts guide has examples.
Asset 6 is the data and dissemination plan in the new Research.gov tool, covering preprints, any code or formal proofs, and how author accepted manuscripts will reach the Public Access Repository.
5 Route Choices Before Drafting
NSF 26-520 covers 14 programs in the mathematical sciences, so the route choice is now a dropdown in one solicitation rather than a choice between documents. Five routes cover most analysis-adjacent work.
| Program | Choose it when | What NSF's description covers |
|---|---|---|
| Analysis | Analysis is the main contribution | Harmonic, complex, real, and functional analysis; dynamics and ergodic theory; operator algebras; PDE; calculus of variations |
| Geometric Analysis | Geometry and analysis jointly carry the argument | Differential geometry, geometric group theory, geometric PDE |
| Probability | Randomness carries the argument | Stochastic processes, interacting particle systems, stochastic PDE, Markov processes |
| Applied or Computational Mathematics | A model or algorithm is the main output | Problems from science and engineering; numerical algorithm questions |
| CAREER or a conference proposal | The ask is a 5-year early-career plan or event support | CAREER runs on NSF 22-586; conferences use the conference proposal type in NSF 26-520 |
The boundaries overlap. Geometric partial differential equations are named in the Geometric Analysis description, and stochastic partial differential equations sit in Probability. If the core estimate is deterministic and analytic, Analysis is the better route; if curvature or randomness carries the argument, compare the NSF Geometric Analysis guide and the NSF Probability guide first.
6 Bad-Fit Signals
Six signals mean you should skip Analysis or rework the draft before submitting. With 88 awards carrying the Analysis code in FY2026, down from 117 the year before, a proposal that belongs in another program is a waste of a submission.
First, analysis is only a tool. Second, geometry or probability dominates the main theorems. Third, the project is mainly numerical method development.
Fourth, the work is an application project with limited analysis contribution. Fifth, the draft still follows PD 20-1281 instructions instead of NSF 26-520. Sixth, the broader impacts plan is a generic training sentence, or it limits participation in ways NSF's current priorities FAQ says it won't fund.
Any one of these is a red flag. Two or more usually mean a better route exists.
Before You Submit: 5 Final Checks
Run these 5 checks before an anytime submission.
- Program and solicitation. NSF 26-520 is selected with Analysis as the program, and nothing in the proposal cites PD 20-1281.
- Budget. The request is in line with the $200,000 to $300,000 band, uses your negotiated indirect cost rate, and includes no voluntary committed cost sharing.
- Timing. You've asked Analysis@nsf.gov how anytime proposals are being reviewed if the start date matters for a student or postdoc.
- PAPPG compliance. The proposal follows NSF 24-1 as modified by Supplements 1 and 2, including the Research.gov data plan tool. The NSF data management plan guide walks through it.
- Route proof. The project description explains in 2 sentences why Geometric Analysis, Probability, Applied Mathematics, or Computational Mathematics is not the better home.
A proposal that fails check 5 isn't worth sending yet. Email the program directors with a one-page summary and ask which program should review it; a short answer from them can save a full rewrite. For how panels and ad hoc reviewers weigh the two criteria, see how NSF grant review works.
Sources
- NSF Analysis (opens in a new tab)
- NSF 26-520: MPS Mathematical Sciences Research Programs (opens in a new tab)
- NSF Proposal and Award Policies and Procedures Guide (opens in a new tab)
- PAPPG 24-1 Supplement 1 (NSF 26-200) (opens in a new tab)
- PAPPG 24-1 Supplement 2 (NSF 26-202) (opens in a new tab)
- NSF updates on priorities (broader impacts and indirect cost FAQ) (opens in a new tab)
- Executive Order 14332, Improving Oversight of Federal Grantmaking (opens in a new tab)
- NSF CAREER solicitation NSF 22-586 (opens in a new tab)
- Grants.gov listing 325500 (opens in a new tab)
- Civil Rights Litigation Clearinghouse: AAU v. NSF (opens in a new tab)
- NSF Award Search API records for program element 128100, awards dated October 1, 2023 through September 30, 2026, pulled October 18, 2026
Largest awards active in FY2024
One row per recipient, its largest award | 5,546 Assistance Listing records | Avg: $1.4M
Association of Universities for Research in Astronomy, Inc.
DC
Associated Universities Inc
VA
Use the visible recipient rows as free source context
This program has no reviewed paid report today.
No reviewed paid report is currently offered for this program. The visible source rows remain free.
Frequently Asked Questions
What does NSF Analysis support?
It supports research in complex, harmonic, and real analysis; dynamical systems and ergodic theory; functional analysis; mathematical physics; operator theory and operator algebras; partial differential equations; and calculus of variations. NSF says the list isn't closed, and other analysis topics will be considered.
What funding opportunity applies now?
Since August 17, 2026, Analysis proposals go through NSF 26-520, MPS Mathematical Sciences Research Programs (MPS MathSci). The older program description, PD 20-1281, had a September 16 to 30 window; its Grants.gov listing closed September 30, 2026.
Is there still a September deadline?
No. NSF 26-520 says proposals are accepted anytime, due by 5 p.m. submitting organization's local time. There's no limit on proposals per organization or per PI under the solicitation.
Who can apply?
Accredited two- and four-year US colleges and universities acting for their faculty, US non-profit non-academic research organizations such as museums, observatories, and professional societies, and federally recognized tribes. NSF 26-520 sets no restriction on who may serve as PI.
How big is a typical Analysis research award?
Grantsights analysis of NSF award records found a median of $200,000 across 55 research awards dated in FY2026, down from $291,310 in FY2024. Most ran 3 years, and only 4 were above $300,000.
Does the 15% indirect cost cap apply to NSF Analysis proposals?
Not right now. NSF says a federal court vacated its 15% indirect cost policy (NSF 25-034) and it isn't in effect, so proposers can use their federally negotiated rate. New awards carry a term that would apply the cap if NSF is later allowed to, and NSF 26-520 says NSF may prefer institutions with lower indirect cost rates when proposals are otherwise equal after merit review.
How should conference proposals be handled?
Submit them under NSF 26-520 with Analysis as the program, at least 6 months before the event for requests up to $50,000 and at least 9 months before for larger requests. Meetings that span several math programs go to the Mathematical Sciences Infrastructure program.
Who should skip this route?
Skip it when Geometric Analysis, Probability, Applied Mathematics, Computational Mathematics, Statistics, Topology, or another NSF program is the better review home, or when you're an early-career faculty member whose better first move is the separate CAREER competition.
Last updated October 18, 2026. Corrected October 18, 2026: what changed. Published by Grantsights from public sources. See our editorial policy, how we build and check guides, corrections and how we make money.
Related Guides
Complete Guide
NSF Grants 2026: Programs, Deadlines, and PAPPG
NSF grants 2026 include 12 new research opportunities worth over $1.5 billion. Compare the October GRFP route, PAPPG rules, and proposal fit.
NSF Geometric Analysis Guide
NSF Geometric Analysis guide: fit under NSF 26-520, anytime submission, 6- and 9-month conference lead times, award sizes, and route checks.
NSF Archaeology Senior Research Awards Eligibility
NSF Archaeology Senior Research Awards eligibility guide for NSF 23-566, source status, anthropological value, access proof, and route choice.
NSF AAPF Guide
NSF AAPF guide: $110,000 a year for 3 years, 5-year PhD window, citizenship rule, the Aug. 15-Oct. 15, 2027 window, 10 fellows in 2026, and host rules.
NSF CEDAR Upper Atmosphere Eligibility Guide
NSF CEDAR guide: NSF 25-510 is archived. Upper-atmosphere proposals now go to GEO Core (NSF 26-516), open anytime, with 200-350 awards expected.
NSF CICI Eligibility: Cyberinfrastructure Security
NSF CICI guide: Cybersecurity Innovation for Cyberinfrastructure eligibility, four program areas, Jan. 20 deadline, budget caps, and route checks.