National Science Foundation (NSF)

The National Science Foundation (NSF) — established by the National Science Foundation Act of 1950 (42 U.S.C. §§ 1861–1887) and further authorized by the CHIPS and Science Act of 2022 — is the primary federal agency funding basic scientific research across all fields except life sciences (which fall under NIH), investing approximately $9 billion/year through competitive peer-reviewed grants to universities, research institutions, and scientists across every state. NSF funds the foundational research that underlies American technological leadership: it funded the internet's precursors (ARPANET protocols), GPS mathematical foundations, Google's original search algorithm, and countless other innovations that generated trillions in economic value long after the initial grant. The CHIPS and Science Act authorized a dramatic NSF expansion — to $19.5 billion by FY2027 — focused on new Technology, Innovation, and Partnerships (TIP) directorates targeting translational research in semiconductors, AI, quantum computing, and clean energy. However, Congress has appropriated well below those authorization levels, and the Trump administration's 2025 budget proposals and DOGE-driven cuts have created significant uncertainty about NSF's funding trajectory. NSF funds approximately 44,000 grants/year supporting roughly 300,000 researchers and graduate students — making it a primary pipeline for the STEM workforce. The agency also funds science, technology, engineering, and math education programs at all levels, and operates the U.S. Antarctic Program and other unique national research facilities.

Current Law (2026)

Legal Authority

• 42 U.S.C. § 1862 — Functions of the Foundation (initiate and support basic scientific research and education in the sciences and engineering; award grants and contracts; foster interchange of scientific information; maintain a register of scientific/technical personnel)
• 42 U.S.C. § 1863 — National Science Board (25 members appointed by the President; sets policies for NSF; approves large awards; publishes the biennial Science and Engineering Indicators report)
• 42 U.S.C. § 1870 — General authority (Director manages the Foundation; authority to award grants, enter contracts, and cooperate with foreign countries; merit review process)
• 42 U.S.C. §§ 18901-19351 — CHIPS and Science Act NSF provisions (established the Technology, Innovation, and Partnerships (TIP) Directorate; authorized significant budget increases; regional innovation engines; translation of research to application)

How It Works

The National Science Foundation is the primary federal agency supporting fundamental research and education in science and engineering — funding approximately 25% of all federally supported basic research at U.S. colleges and universities. Unlike mission agencies (NIH funds health research, DOE funds energy research — see DOE high-end computing for one major example), NSF supports research across all fields of science and engineering based on scientific merit.

NSF funds research and education across virtually every field of science, engineering, and mathematics — from astrophysics to zoology, computer science to sociology — through seven directorates: Biological Sciences, Computer and Information Science and Engineering, Engineering, Geosciences, Mathematical and Physical Sciences, Social/Behavioral/Economic Sciences, and the newly created Technology, Innovation, and Partnerships (TIP) directorate. Both investigator-initiated ("curiosity-driven") research and targeted national-priority programs qualify. NSF-funded research has produced breakthroughs including the internet (ARPANET/NSFNet), MRI technology, GPS improvements, and the detection of gravitational waves; when these produce patentable inventions, the Bayh-Dole Act allows universities to retain ownership and license them commercially. NSF's peer review process is considered the gold standard for scientific funding — proposals evaluated by external expert panels on intellectual merit (potential to advance knowledge) and broader impacts (potential to benefit society). The ~25% funding rate makes competition intense; most rejected proposals are excellent but simply unfunded. This process ensures funding decisions rest on scientific quality rather than political factors.

NSF also funds STEM education at every level: K-12 teacher preparation, undergraduate Research Experiences for Undergraduates (REU), the NSF Graduate Research Fellowship ($37,000/year for 3 years), postdoctoral support, and broadening-participation programs for underrepresented groups. Major research infrastructure — telescopes, research vessels, computing centers, and the U.S. Antarctic Program — rounds out NSF's portfolio. The CHIPS and Science Act (2022) created NSF's first new directorate in decades: Technology, Innovation, and Partnerships (TIP), focused on translating basic research into real-world applications, supporting use-inspired research, and strengthening regional innovation ecosystems through "Regional Innovation Engines" — a significant shift toward applied research and economic competitiveness alongside NSF's traditional basic research mission.

How It Affects You

If you're a researcher applying for NSF funding: NSF receives roughly 50,000 proposals per year and funds about 12,000 — a ~25% success rate that makes competition intense but not hopeless. Proposals are evaluated on two criteria: intellectual merit (does this advance scientific knowledge?) and broader impacts (does this benefit society, train future scientists, broaden participation?). The broader impacts criterion is not optional and is weighted equally — reviewers will kill a technically excellent proposal that has a thin broader impacts statement. Proposals are accepted year-round for most programs, but major programs have specific target dates (listed at nsf.gov/awards/managing/proposal_faqs.jsp). Before submitting, read the specific program solicitation — not just the directorate overview. NSF's FastLane/Research.gov submission system requires institutional sign-off through your sponsored programs office, so build in 2-3 weeks for institutional review. Rejection decisions include reviewer comments; these are unusually transparent compared to other federal funders and worth using for revision.

If you're a graduate student in STEM: The NSF Graduate Research Fellowship Program (GRFP) is one of the best funding opportunities in American science. It pays a $37,000/year stipend plus $16,000/year cost-of-education allowance for 3 years — more than most university teaching assistantships, with no teaching requirement. You must apply in your first or second year of graduate study (or as a senior undergrad); post-second-year applicants are permanently ineligible. The fellowship follows you if you transfer institutions and can be deferred for up to two years. About 2,000 fellowships are awarded annually from ~12,000 applications. The application requires two personal statements (personal, relevant background and future goals + graduate research plan) plus three reference letters. Many successful applicants have done undergraduate research, have a clear research agenda, and can articulate how their work connects to NSF's national priorities (AI, climate, quantum, bioeconomy). Check your institution's GRFP advising office — schools with strong GRFP track records often have writing workshops and can share successful past essays.

If you work at a research university or lead a STEM department: NSF is typically among the top 2-3 external funding sources for major research universities, providing not just direct research costs but indirect cost recovery (facilities and administrative costs, typically 50-60% of direct costs) that funds university infrastructure. The CHIPS and Science Act (2022) authorized NSF to grow from ~$8.8 billion to $19.9 billion over five years — though actual Congressional appropriations have significantly lagged authorizations, leaving a funding gap. The new TIP directorate is specifically designed to fund translational research and regional innovation, with grants often larger than traditional research awards. NSF's EPSCoR program (Established Program to Stimulate Competitive Research) targets 28 states, Puerto Rico, Guam, and the Virgin Islands that have historically received less federal research funding — check whether your state qualifies for EPSCoR supplements that can add 10-20% to award values. Track NSF program officer turnovers at nsf.gov — program officers are accessible by email and will give substantive feedback on whether a proposal idea fits a program before you spend 200 hours writing it.

If you're watching federal research funding as a policy matter: NSF's ~$9.9 billion annual budget funds roughly 25% of all federally supported basic research at U.S. universities — the kind of foundational science that doesn't have immediate commercial applications but underpins future technology. The internet traces back to NSFNet. MRI technology to NSF-funded materials science. Gravitational wave detection to decades of NSF investment. The return on basic research investment is real but long-delayed — which makes it politically vulnerable. The Trump administration's 2025 budget proposed cutting NSF by ~15%, and the DOGE-era scrutiny of grants labeled "woke" or "ideological" led to termination of some social science and DEI-related awards in early 2026. Research security requirements — especially disclosures of ties to China and other "countries of concern" — have increased compliance burdens on NSF-funded researchers, with some universities pulling back from international collaborations. The NSF director is a presidential appointee confirmed by the Senate; changes in leadership shift priorities across the agency's 7 directorates.

State Variations

NSF is exclusively federal. However, NSF-funded EPSCoR (Established Program to Stimulate Competitive Research) specifically targets states that historically receive less federal research funding, building research capacity in underserved states.

Implementing Regulations

• 2 CFR Part 2500 — NSF Uniform Administrative Requirements (supplement to Uniform Guidance for NSF grants)

• 45 CFR Part 600–690 — NSF grant and agreement regulations (proposal requirements, review procedures, award conditions)

• 45 CFR Part 689 — Research Misconduct: NSF's regulations for investigating and sanctioning fabrication, falsification, and plagiarism (FFP) in NSF-funded research — the three practices that constitute research misconduct under federal policy:

  • § 689.1 — Definition of research misconduct: NSF limits research misconduct to three specific practices — fabrication (inventing data or results and reporting them), falsification (manipulating research materials, equipment, images, or processes; or changing or omitting data so that the record is not accurately represented), and plagiarism (appropriating another person's ideas, processes, results, or words without giving appropriate credit); the definition explicitly excludes honest error and differences of scientific opinion — contested interpretations and failed replications are not misconduct
  • § 689.3 — Sanctions (three tiers): NSF imposes a tiered sanction system: Group I (least severe) includes letters of reprimand, additional certification or assurance requirements, or required supervision; Group II includes award restrictions, special review conditions for new proposals, or restrictions on serving as reviewer or advisory committee member for a period of 1–3 years; Group III (most severe) includes debarment from participation in NSF programs for a period of 5+ years, termination of active awards, and referral to appropriate law enforcement agencies; sanctions are calibrated to the seriousness, extent, and intent of the misconduct
  • § 689.4 — Awardee institution primary responsibility: universities, research institutions, and other NSF-funded entities bear primary responsibility for preventing and detecting research misconduct and for conducting inquiries and investigations of allegations; NSF typically defers to the institution's investigation results and relies on the institution's process before imposing sanctions; institutions must report findings to NSF's Office of Inspector General (OIG) within 30 days of completing an investigation that finds misconduct
  • § 689.5 — NSF OIG handling: NSF staff who learn of alleged misconduct must report to OIG; the identity of informants who wish to remain anonymous is protected to the extent permitted by law; OIG decides whether to open an inquiry or investigation or to defer pending the institution's investigation
  • § 689.6 — NSF investigations: when the institution's investigation is underway, OIG may defer its own inquiry; if OIG doesn't receive institutional results within 180 days, it may proceed independently; NSF investigations may examine whether the institutional process was adequate and whether the institution's findings should be adopted, modified, or rejected
  • § 689.8 — Interim actions: during an investigation (before a final finding), NSF may take interim protective measures — temporarily restricting an individual from serving as a PI, reviewer, or advisory committee member; placing special award conditions on an institution; or deferring action on pending proposals; interim actions require "reasonable belief" that misconduct has occurred and are taken to protect federal resources, not as punishment
  • § 689.10 — Appeals: an individual or institution may appeal to the NSF Director within 30 days of receiving the Deputy Director's written decision; after the Director's ruling, judicial review is available through the federal courts

  NSF's research misconduct framework implements the government-wide Federal Policy on Research Misconduct (OSTP, December 2000), which established FFP as the federal definition and required each agency to implement its own procedures. NSF's Part 689 is one of the more comprehensive agency implementations. The research misconduct investigation system operates primarily through institutions because they have direct access to laboratory records, notebooks, data files, and research personnel — NSF typically cannot investigate these materials without institutional cooperation. Cases that reach NSF OIG enforcement typically involve finding fabricated data in published papers, plagiarism in grant applications, or systemic data manipulation affecting multiple funded projects. Recent rulemakings: 67 FR 11937 (March 2002) — original Part 689 rule; 72 FR 4944 (February 2007) — clarifying amendments.

Pending Legislation (119th Congress)

• SRES 508 — Designates November 8, 2025 as "National STEM Day" to highlight STEM education and workforce needs and encourage programs promoting STEM careers. Status: Passed Senate.

Recent Developments

• DOGE-driven grant terminations (2025): NSF faced some of the most dramatic DOGE interventions in the federal science funding landscape. Hundreds of NSF grants were terminated or suspended for containing DEI-related research, social science research on topics deemed politically sensitive, or climate-related research. The mass terminations affected projects in social sciences, education research, gender studies, environmental science, and fields perceived as advancing a progressive agenda. Researchers whose grants were terminated challenged the terminations as arbitrary and capricious under the APA; courts issued some injunctions staying terminations pending review.
• CHIPS and Science Act implementation lagging: The CHIPS and Science Act (2022) authorized NSF to grow from ~$9 billion to ~$18 billion annually and created the new Technology, Innovation and Partnerships (TIP) Directorate for applied research. Actual appropriations have significantly lagged authorizations — Congress never fully funded the CHIPS science provisions. The Trump administration's deregulatory posture and budget constraints further reduced the likelihood of the authorized funding levels being appropriated.
• Research security and China: NSF's research security requirements — including disclosure of foreign research affiliations and participation in foreign talent recruitment programs — have been enforced more aggressively alongside export control tightening and cybersecurity research workforce standards. DOJ prosecutions of researchers who failed to disclose Chinese government affiliations while receiving NSF grants continued, though the prosecution rate moderated from its peak. The tension between openness (essential to science) and security (essential to protecting advanced research) remains a defining challenge for NSF.
• AI research priority: NSF continued prioritizing AI fundamental research — supporting the National AI Research Institutes and AI safety research programs. The Trump administration's AI EO supported AI research investment, making NSF's AI programs less politically contested than its social science or climate programs — see Artificial Intelligence Policy and the related National Quantum Initiative for NSF's role in cross-cutting emerging-technology programs.