The National Science Foundation (NSF) is an independent US federal agency and the primary public funder of basic research and STEM education outside biomedical and clinical science. Where the National Institutes of Health funds research into human health and disease, NSF’s statutory mandate covers “all fields of fundamental science and engineering, except for the medical sciences” — mathematics, physics, chemistry, computer science, engineering, the geosciences, the social and behavioral sciences, and STEM education itself. Per NSF’s own reporting, its roughly $8.75 billion FY2026 budget accounts for about 25% of all federal support for basic research at US colleges and universities, funding around 11,000 new competitive awards a year across some 1,900 institutions and reaching approximately 350,000 researchers, students, teachers, and entrepreneurs.
For a research administrator, investigator, or anyone new to sponsored programs, “NSF funding” is not one undifferentiated pool of money — it is a structured system of award instruments, organizational directorates, and a distinctive two-criterion review process that differs in real, operational ways from NIH and other federal funders. This guide is the orientation page for that system: what NSF funding is, its main award-instrument types, how NSF’s directorate structure organizes programs, how proposals are reviewed, and where to find current opportunities. It deliberately does not re-derive detail already covered on CASRAI’s mechanism-specific pages — those are linked throughout — so treat this as the map, not the full territory.
What NSF funding actually is
NSF was established by the National Science Foundation Act of 1950 with a mission built around three goals: promote the progress of science, advance the national health, prosperity, and welfare, and secure the national defense. In practice, NSF pursues that mission almost entirely through extramural funding — supporting investigator-initiated and program-solicited research and education activities at universities, colleges, and other organizations, rather than running large in-house laboratories the way some other federal science agencies do.
NSF funds more than conventional research projects. Alongside standard research grants, it supports individual and institutional fellowships and traineeships (including its long-running Graduate Research Fellowship Program), major research infrastructure and multi-user facilities, education and workforce-development programs, and, since the CHIPS and Science Act of 2022, technology-translation and use-inspired research programs housed in a newer directorate (see below). Because NSF, like every other federal awarding agency, is bound by the government-wide Uniform Guidance, the cost principles, audit requirements, and administrative rules that apply to an NSF award are largely the same framework that governs NIH, DOE, and other federal research funding — see CASRAI’s Uniform Guidance (2 CFR 200) guide for that shared baseline, and Federal Government Contracts vs. Grants for University Research for how NSF’s grant-based approach differs from a procurement contract.
NSF’s main funding-instrument types
NSF’s Proposal & Award Policies & Procedures Guide (PAPPG) — currently NSF 24-1, with periodic policy-notice supplements — defines the award instruments NSF uses. As with NIH, the differences are not just paperwork; they determine how much involvement NSF retains in a project after award:
- Standard grants provide a specific, fixed level of support for a specified period, with no NSF commitment to future funding beyond that period. A new proposal (a renewal) is required to seek additional support once the award period ends.
- Continuing grants provide support for an initial period — typically one year — together with a statement of NSF’s intent to fund additional increments in future years, provided that funds are available and progress warrants it. Continuing-grant increments are treated as a high priority within NSF’s budget process and generally are not put back into competition with new proposals, but they are not a guarantee independent of performance and appropriations.
- Cooperative agreements are used instead of a grant when NSF anticipates “substantial involvement” with the recipient in carrying out the funded activity — joint planning, milestone-based oversight, or direct NSF participation in the work, rather than a purely investigator-initiated project. NSF’s PAPPG flags this instrument for large research centers, multi-user facilities, projects with complex subcontracting structures, and projects where NSF is one of several stakeholder organizations with real influence over direction. Large-facility construction cooperative agreements are funded separately through NSF’s Major Research Equipment and Facilities Construction (MREFC) account. CASRAI’s Grant vs. Contract vs. Cooperative Agreement comparison covers the underlying legal distinction (rooted in the Federal Grant and Cooperative Agreement Act of 1977, 31 U.S.C. §§ 6303–6305) in more depth.
NSF has also used Other Transactions (OT) authority — a statutory authority separate from standard grants, cooperative agreements, and procurement contracts, granted to NSF by the CHIPS and Science Act of 2022 — for a small number of milestone-based programs; see CASRAI’s guide to NSF X-Labs and Other Transactions (OT) Authority for how that mechanism differs from the three instruments above. Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) awards are a related but distinct category: they operate under a separate statutory framework (the SBA’s SBIR/STTR Policy Directive) that NSF implements through its own competitions, run administratively as grants but with their own phase structure and eligibility rules.
How NSF’s directorate structure works
NSF organizes its research and education programs into directorates, each covering a broad disciplinary or functional area and led by an Assistant Director. As of this writing, following NSF’s recent internal reorganization, the directorates are:
- Biological Sciences (BIO)
- Computer and Information Science and Engineering (CISE)
- Engineering (ENG)
- Geosciences (GEO)
- Mathematical and Physical Sciences (MPS)
- Social, Behavioral and Economic Sciences (SBE)
- STEM Education (EDU)
- Technology, Innovation and Partnerships (TIP)
The Office of the Director and the National Science Board — NSF’s presidentially-appointed policy-setting body — sit above the directorates, along with administrative offices (Budget, Financial Officer, Information Officer, Inspector General, and others) that operate across all of them. Each directorate is further divided into divisions and programs, each with an assigned program officer who is usually a scientist or engineer on temporary rotation from academia or industry (or, in some cases, permanent NSF staff) and is typically the most useful point of contact when scoping a proposal to a specific solicitation.
TIP, established under the CHIPS and Science Act of 2022, is NSF’s newest directorate and is organizationally distinct from the others in that its programs are explicitly oriented toward accelerating use-inspired and translational research toward societal and economic impact, rather than curiosity-driven basic research alone — NSF’s I-Corps entrepreneurial-training program and much of NSF’s Other Transactions activity sit within it. See CASRAI’s NSF I-Corps Program guide for how one TIP-housed program actually runs. Award terms, eligibility details, and even proposal-preparation conventions can vary meaningfully by directorate — NSF’s own CAREER program, for example, sets a higher minimum award value for proposals in BIO, ENG, and the Office of Polar Programs than for other directorates — so a research administrator should always confirm directorate- or program-specific guidance in the relevant funding opportunity rather than assuming uniform rules apply NSF-wide.
How NSF reviews proposals: the merit review criteria
Every proposal NSF accepts for review is evaluated against two National Science Board-approved merit review criteria, applied together rather than as alternatives:
- Intellectual Merit — the potential for the proposed work to advance knowledge and understanding within its field or across fields.
- Broader Impacts — the potential for the proposed work to benefit society and contribute to specific, desired societal outcomes: for example, improving STEM education, broadening participation in STEM, building academia–industry partnerships, strengthening national security, or improving research and education infrastructure.
NSF revised how it expects Broader Impacts to be evidenced starting in 2024, with rolling implementation through 2025–2026: reviewers are now expected to weigh plans grounded in the broader-impacts and informal-STEM-learning research literature, with measurable, observable outcomes, over generic or merely plausible-sounding activities. CASRAI’s news coverage of that revision, What “broader impacts” means under the new NSF policy, goes into what changed and what it means for applicants preparing a Broader Impacts plan today.
Mechanically, an NSF program officer first checks a proposal against basic eligibility and compliance requirements before it goes to review. Substantive review is then typically conducted by a panel of external subject-matter experts, by mail (ad hoc) reviewers, or a combination of both, convened by the program officer. Panelists individually score and write reviews against both criteria, then discuss and rank proposals as a panel. The program officer weighs that reviewer input alongside program-level considerations — portfolio balance across research approaches, potential for transformative results, and specific program objectives — when recommending an award; a strong review does not by itself guarantee funding, and NSF’s overall funding (award) rate across all directorates typically runs in the mid-to-high twenties as a percentage of proposals received. NSF’s stated target is to notify applicants of a funding decision within six months of the proposal deadline.
Where to find current NSF funding opportunities
NSF posts every active funding opportunity on its NSF Funding Search, searchable by keyword and filterable by directorate, division, and other criteria; a directorate-first browsing view is also available directly. Two other NSF systems are worth knowing alongside the opportunity search:
- Research.gov is NSF’s grants-management platform for the research community — where most proposals are actually prepared and submitted, project reports are filed, and (through its Award Search) prior NSF awards dating back to 1989 can be searched by title, investigator, institution, or program, which is useful for scoping how competitive a given program area has historically been.
- The PAPPG (NSF 24-1) is the single governing document for how NSF proposals must be prepared, submitted, reviewed, and administered once awarded — the NSF equivalent of a combined solicitation-format guide and post-award administration manual. It is revised periodically through numbered policy notices and supplements rather than a full annual reissue, so always confirm which PAPPG version and supplement applies to a given proposal’s due date before relying on chapter-and-section detail from memory or an older copy.
Individual funding opportunities are published as program solicitations (analogous to NIH’s Notices of Funding Opportunity) or, for unsolicited proposals falling within a program’s general scope, submitted at any time under that program’s standing description in the PAPPG. Deadlines, award sizes, and submission windows vary program-by-program and directorate-by-directorate — there is no single NSF-wide deadline calendar the way NIH maintains standard due dates for some mechanisms — so the specific solicitation, not this overview, is always the controlling document for eligibility, page limits, and deadlines.
Administrative touchpoints research administrators should know
A few compliance and administration requirements are distinctive enough to NSF, versus other federal funders, that they are worth flagging even in an orientation guide:
- Responsible and ethical conduct of research (RECR) training is a statutory NSF requirement under Section 7009 of the America COMPETES Act of 2007 (as later amended, including by the CHIPS and Science Act of 2022), distinct from NIH’s RCR expectations, which arise from grants policy rather than a dedicated statute. NSF requires institutions to certify, at proposal submission, that they have a plan to train undergraduates, graduate students, and postdoctoral researchers supported on the award. See CASRAI’s Responsible Conduct of Research (RCR) Training guide and the RCR training dictionary entry for how that obligation is generally administered.
- Data management plans are a required part of every NSF proposal, but NSF’s requirements and NIH’s 2023 Data Management and Sharing Policy differ in real, not just cosmetic, ways — NSF tailors its expectations by directorate, and, for proposals submitted on or after April 27, 2026, requires the plan through a structured Research.gov webform rather than an uploaded PDF. CASRAI’s NIH vs. NSF Data Management Plans guide walks through what actually differs.
- Indirect costs (facilities and administrative, or F&A, costs) on NSF awards follow each institution’s federally negotiated rate under the same Uniform Guidance framework NIH and other agencies use — see CASRAI’s Indirect Cost Rate (F&A Rate) entry.
- CRediT authorship statements for manuscripts arising from NSF-funded research follow the journal’s own policy in the first instance, but CASRAI maintains funder-specific guidance on NSF’s expectations at NSF and CRediT statements for investigators preparing publications that need to credit NSF support and individual contributions correctly.
- Most institutions split administration of an NSF award, like other federal awards, between a pre-award function (proposal development, budgeting, submission) and a post-award function (award setup, financial monitoring, reporting, closeout) — see CASRAI’s Pre-Award vs. Post-Award Office Roles guide for how that split typically works, and What a Research Administrator Does for the role that sits across both.
Frequently asked questions
What is the difference between an NSF grant and an NSF cooperative agreement?
A standard or continuing NSF grant involves limited NSF programmatic involvement once the award is made. A cooperative agreement is used instead when NSF expects “substantial involvement” in carrying out the funded activity — joint planning, milestone oversight, or direct participation — typically for large centers, shared research infrastructure, or projects with complex multi-organization structures.
What are NSF’s two merit review criteria?
Intellectual Merit (the potential to advance knowledge within or across fields) and Broader Impacts (the potential to benefit society through outcomes such as STEM education, broadened participation, or infrastructure). Both are applied to every proposal; neither is scored in isolation from program-level funding decisions.
How many directorates does NSF have?
Eight, as of this writing: Biological Sciences (BIO), Computer and Information Science and Engineering (CISE), Engineering (ENG), Geosciences (GEO), Mathematical and Physical Sciences (MPS), Social, Behavioral and Economic Sciences (SBE), STEM Education (EDU), and Technology, Innovation and Partnerships (TIP). NSF periodically adjusts its internal structure, so confirm current directorate names against nsf.gov before citing them in a proposal.
What is the PAPPG?
The Proposal & Award Policies & Procedures Guide — NSF’s single governing document for how proposals must be prepared and submitted and how resulting awards are administered. The current version is NSF 24-1, updated periodically through numbered policy notices and supplements rather than a full annual reissue.
Is NSF the same as NIH?
No. Both are US federal science funders, but NIH funds biomedical and health-related research through 27 Institutes and Centers, while NSF funds basic research and STEM education across all other fields of science and engineering through its directorate structure. Their award mechanisms, review processes, and policy requirements (data management plans and RCR training among them) diverge in specific, operationally relevant ways — see CASRAI’s NIH vs NSF vs ERC comparison for one concrete example.
Where do I find NSF grant deadlines?
In the specific program solicitation on NSF’s Funding Search — NSF does not maintain a single agency-wide deadline calendar the way NIH publishes standard due dates for some mechanisms. Deadlines, target dates, and “any time” submission windows are set program-by-program.
Related CASRAI resources
- Grants Management & Research Funding — the cluster hub for grant mechanisms, compliance, and international funding-body content
- NIH Grants: An Overview — the equivalent orientation guide for NIH, NSF’s closest federal-funder counterpart
- Uniform Guidance (2 CFR 200)
- Grant vs. Contract vs. Cooperative Agreement
- NSF I-Corps Program
- NSF X-Labs and Other Transactions (OT) Authority
- NIH vs. NSF Data Management Plans
- What “broader impacts” means under the new NSF policy
- NSF Research Security: What Notice 149 Requires Proposers to Certify
