SBIR & Research
NSF Grants: Programs, Eligibility, and How to Apply
Grant Writing Service Editorial Team
June 17, 2026 · 5 min read
Table of contents
Key takeaways
- NSF funds non-medical science, engineering, and STEM education research, mostly through institutions that submit on a principal investigator's behalf.
- Every proposal is reviewed against Intellectual Merit and Broader Impacts; the small-business program adds a third criterion for commercial impact.
- Your real target is a specific program and its solicitation, not the agency as a whole.
- NSF replaced its divisions with smaller sections in late 2025 and changed some review practices, so read the current solicitation rather than an older one.
An NSF grant is funding from the U.S. National Science Foundation, the independent federal agency that supports fundamental research and education across science and engineering, excluding medical research, which belongs at NIH. Awards are made through competitive programs, most proposals are submitted by an institution on behalf of a principal investigator, and every proposal is judged on two merit review criteria: Intellectual Merit and Broader Impacts.
How NSF is organized now
NSF funds physics, chemistry, biology, computer science, engineering, the geosciences, the social and behavioral sciences, and STEM education through research directorates. In December 2025 NSF kept its directorates but replaced the divisions inside them with smaller units called sections, and some directorates reorganized further in early 2026. Program names, contacts, and funding lines shifted in the process, which is one more reason to work from the current solicitation rather than an older one or a colleague's past proposal.
What did not change is that the program is your real target. Each program has a program officer who manages its portfolio, runs its review, and recommends awards. A short email to that officer with a one-paragraph summary of your project, asking whether it fits, is standard practice and often the most useful early step.
NSF also issued a Statement of Priorities in April 2025 that still shapes how programs read proposals. It ties broader impacts to the goals in the America COMPETES Reauthorization Act of 2010 and says activities to broaden participation must create opportunities open to all Americans. Proposals that preference or exclude participants by protected characteristics do not fit current policy.
Program types
NSF runs hundreds of funding lines, but most applicants are choosing among a few kinds:
- Standard and continuing research grants submitted to a disciplinary program, typically for about three years.
- The CAREER program, NSF's early-career award for faculty who integrate research and education.
- The Graduate Research Fellowship Program, which funds individual students rather than projects. Its eligibility and eligible fields have narrowed in recent cycles, so read the current solicitation before assuming you qualify.
- Education and infrastructure programs, including research experiences for undergraduates and equipment awards.
- America's Seed Fund, NSF's SBIR and STTR program for small businesses developing deep technology, covered below.
How NSF differs from NIH
Researchers who know one agency often assume the other works the same way. It does not:
| NSF | NIH | |
|---|---|---|
| Scope | Basic science, engineering, education | Biomedical and health research |
| Review criteria | Intellectual Merit and Broader Impacts | Importance of the research, rigor and feasibility, plus an expertise and resources check |
| Core narrative | Project description, 15 pages for most proposals | Specific Aims (1 page) plus Research Strategy |
| Who runs review | Program officer convenes a panel and may add ad hoc reviewers | Center for Scientific Review study section, then institute council |
| Submission | Research.gov, or Grants.gov where the solicitation allows | Grants.gov through eRA Commons |
An NIH Specific Aims page does not map onto an NSF project summary, so expect to rebuild the proposal's architecture when moving between them. Our guide to NIH grants covers the NIH side.
What reviewers are asked
Intellectual Merit is the potential to advance knowledge. Reviewers look for an important question, a sound and creative plan, and a team and setting able to carry it out.
Broader Impacts is the potential to benefit society and contribute to specific outcomes, such as training, public understanding, infrastructure others can use, or application of the results. It is scored as seriously as the research, and it is where otherwise strong proposals most often lose ground.
For both criteria, reviewers consider whether the activities are well reasoned and organized, whether there is a mechanism to assess success, whether the team is qualified, and whether resources are adequate. The project summary must address each criterion in its own labeled statement, and proposals missing that separation can be returned without review.
In December 2025 NSF also changed some review practice: program officers have more latitude to decide when a panel discussion is needed, and the minimum number of external reviews for many proposals dropped from three to two, with internal review able to substitute for one. The criteria themselves were not rewritten.
Broader Impacts sections that hold up
Weak Broader Impacts sections fail in recognizable ways:
- Generic dissemination. "We will publish and present at conferences" describes normal research, not an impact activity.
- Bolted on. Activities with no budget line, no named partner, and no person responsible read as written the night before submission.
- Unmeasured. No way to tell afterward whether anything happened.
A section that survives panel review names the activity, the partner who has agreed to take part, the budget that pays for it, the person responsible, and how success will be measured. If your activities involve schools, our overview of STEM grants for K-12 helps identify partners, and our budget line item rules cover costing the work so it appears in the budget, not only the narrative.
Who can apply and how to submit
For most research awards the applicant is the institution: a university or eligible organization registered in SAM.gov and Research.gov that submits on the principal investigator's behalf and manages the award. Individuals generally cannot hold a standard research grant without an eligible institutional home, though the fellowship program funds students directly.
The rules live in NSF's Proposal and Award Policies and Procedures Guide. The current base edition is NSF 24-1, effective for proposals due on or after May 20, 2024, updated by supplements issued in December 2025 and January 2026. A full replacement guide was in draft for public comment in 2026, so check which version your solicitation cites. Page limits, required sections (including the data management and sharing plan, and biographical sketches prepared in SciENcv), and formatting are checked, and noncompliant proposals are returned without review. For cross-agency mechanics such as registrations, see applying for federal grants.
Deadlines are program by program: some programs have fixed deadlines, some use windows or target dates, and some accept proposals at any time. Read the dates in the current solicitation and set your internal deadline at least a week earlier to leave time for institutional review.
The small-business route: America's Seed Fund
NSF's SBIR and STTR program funds for-profit small businesses rather than institutions. It requires a short Project Pitch before a full proposal is invited, and its review adds a Commercial Impact criterion alongside Intellectual Merit and Broader Impacts. The current solicitation (NSF 26-510, issued after the programs were reauthorized in April 2026) offers Phase I awards of up to $305,000 and Phase II awards of up to $1,250,000. Our guide to writing an SBIR proposal covers how this differs from NIH and Department of Defense submissions.
If you want reviewers' objections answered before the panel raises them, our NSF grant writing service builds Broader Impacts as a designed component of the proposal, not a closing paragraph.
