Applying for a National Science Foundation (NSF) grant follows a specific procedural sequence that differs in real, load-bearing ways from other US federal sponsors: NSF proposals are prepared and submitted natively through Research.gov (not eRA Commons, which is NIH’s system), governed by NSF’s own Proposal & Award Policies & Procedures Guide (PAPPG) rather than the NIH grants policy statement, and reviewed against two NSF-specific merit-review criteria — Intellectual Merit and Broader Impacts — that don’t map directly onto another agency’s review language. This guide walks through that NSF-specific process end to end: account setup, proposal preparation per the PAPPG, submission, and what happens after you submit.
For the general shape of applying for a research grant from any sponsor — eligibility basics, working with your Sponsored Programs Office, budget development — see CASRAI’s How to Apply for a Research Grant guide. For an orientation to NSF’s funding structure itself (award-instrument types, directorates, where to find opportunities), see NSF Grants: An Overview. This guide assumes you already know you’re applying to NSF specifically and need the procedural steps.
Before you start: eligibility and your Sponsored Programs Office
NSF proposals are submitted by an institution on a researcher’s behalf, not by the researcher directly. Almost every college, university, or nonprofit research organization requires proposals to be routed through and formally submitted by its Sponsored Programs Office, acting through a staff member with delegated authority to bind the institution. Contact that office as early as possible — most institutions have internal deadlines well ahead of NSF’s own, especially for budget review and required institutional certifications (human subjects, export control, conflict of interest).
Confirm two things before investing time in a specific opportunity: that you meet the eligibility criteria stated in the funding opportunity (career stage, institutional affiliation, prior award history for some mechanisms), and that your institution is currently registered to receive federal funds.
Step 1: Register for an NSF ID and set up your Research.gov account
Every individual who will have a role on an NSF proposal — principal investigator (PI), co-PI, or senior personnel — needs an NSF ID and a Research.gov account. Before registering, search the NSF ID Lookup tool to confirm you don’t already have an account from a prior proposal, award, or panel-review role; NSF ID accounts persist across institutional moves and are not reissued. To register: go to Research.gov, select Register, and provide the requested personal information, including both a primary and a work email address. Once registered, proposal preparation and submission tools appear under “Prepare and Submit Proposals” on your Research.gov desktop.
This is a genuinely NSF-specific step: NSF retired FastLane as a proposal-submission option for all funding opportunities effective January 30, 2023 (under PAPPG NSF 23-1), with remaining FastLane functions such as award documents and organizational reports decommissioned in phases afterward. Research.gov is now NSF’s native system for proposal preparation, submission, and post-award reporting. See CASRAI’s eRA Commons vs. Research.gov comparison if you also work with NIH submissions and need the two systems’ conventions side by side.
Step 2: Confirm your institution’s federal registrations
Registration at the institutional level is handled by your Sponsored Programs Office, not by individual investigators, but it’s worth knowing it exists: an institution must maintain an active registration in SAM.gov (System for Award Management) with a Unique Entity Identifier, and must be separately registered with NSF itself, before it can submit or receive an NSF award. If your institution has received federal funding before, this registration almost certainly already exists and just needs to stay current — SAM.gov registrations require annual renewal.
Step 3: Identify the right funding opportunity
NSF organizes funding through both standing “core” programs (recurring solicitations within a division, often with no submission deadline or multiple deadlines a year) and one-off program solicitations, sometimes announced first via a Dear Colleague Letter signaling a funding priority before a formal solicitation follows. Search current opportunities directly on Research.gov or NSF’s program pages, organized by directorate and division — see CASRAI’s NSF Funding Search guide for a walkthrough of the search tools themselves, and NSF Award Search if you want to see what NSF has funded previously in your area before you write. Read the specific solicitation’s eligibility, page limits, and required sections closely — these vary by program and can differ from PAPPG defaults.
Step 4: Prepare the proposal package per the PAPPG
NSF’s PAPPG (current version NSF 24-1, applicable to proposals submitted on or after May 20, 2024, with later policy-notice supplements layered on top) specifies exactly what a complete proposal package contains and how each section must be formatted. The core required elements are:
- Project Summary — limited to one page, and must contain three separately labeled sub-sections: an Overview (the activity’s description, objectives, and methods), a statement addressing Intellectual Merit (the activity’s potential to advance knowledge), and a statement addressing Broader Impacts (its potential to benefit society or achieve specific, desired societal outcomes). These two labels aren’t just formatting — Intellectual Merit and Broader Impacts are NSF’s two official merit-review criteria, applied to every proposal NSF receives regardless of program.
- Project Description — the substantive research narrative, subject to a page limit set by PAPPG (15 pages by default, though specific solicitations sometimes set a different limit) that must also explicitly address both merit-review criteria in the body.
- References Cited, Budget and Budget Justification, Facilities/Equipment/Other Resources, and Biographical Sketches — biosketches must be prepared using NSF-approved formats generated through SciENcv, NSF’s required tool for this document (not a free-text upload).
- Current and Pending (Other) Support — also prepared in SciENcv.
- Collaborators and Other Affiliations (COA) — unlike the two documents above, COA is not prepared in SciENcv; NSF requires its own standard .xlsx spreadsheet template, uploaded directly as a supplementary document in Research.gov.
- Data Management Plan — required for nearly all NSF proposals. As of April 27, 2026, NSF moved this from an uploaded PDF to a structured webform inside Research.gov, with the specific fields tailored to the proposal’s primary directorate (proposals already in progress before that date may still use the PDF format). See CASRAI’s NIH vs. NSF Data Management Plans guide for how NSF’s requirement compares to NIH’s, and the Data Management Plan dictionary entry for the concept generally. Note that a structured webform is not the same thing as a machine-actionable DMP conforming to the RDA DMP Common Standard — don’t conflate the two when advising a PI.
Every section has its own formatting rules (font, margins, page limits) enforced at submission — a proposal that violates them can be returned without review rather than simply flagged. Read PAPPG Chapter II in full for the proposal you’re preparing rather than relying on a prior proposal’s formatting, since NSF revises the PAPPG roughly annually and specific solicitations can layer additional requirements on top.
Step 5: Route the proposal internally for institutional sign-off
Before submission, your Sponsored Programs Office reviews the complete package — budget, compliance certifications, institutional commitments — and formally endorses it through Research.gov on the institution’s behalf; NSF requires this same institutional endorsement structure that every US federal sponsor does. Build in time for this step separately from your own writing deadline; it is a common, avoidable cause of late or non-submitted proposals when treated as an afterthought. See the general How to Apply for a Research Grant guide’s “Internal review and routing” section for how this step typically works across sponsors.
Step 6: Submit through Research.gov
The completed, institutionally-endorsed proposal is submitted natively through Research.gov. Grants.gov remains available as an alternate submission path for some NSF opportunities, but Research.gov is NSF’s primary, purpose-built system and is required for most current solicitations. Submission triggers an automated compliance check for missing sections, page-limit violations, and formatting errors before the proposal is accepted into NSF’s processing queue — resolve any errors it flags immediately, since a proposal that fails this check is not received by NSF’s deadline even if you uploaded it before the deadline passed.
Step 7: What happens after you submit
Once accepted, a Program Officer in the relevant division is assigned and the proposal enters merit review: evaluation by external reviewers (ad hoc mail reviewers, a panel of reviewers convened to discuss and rank a group of proposals together, or both, depending on the program) against the Intellectual Merit and Broader Impacts criteria. The Program Officer weighs the reviews and reviewer discussion, and recommends either declination or an award; a Division Director must concur before an award recommendation moves forward, and NSF’s Division of Grants and Agreements (the grants officials with actual authority to obligate federal funds) issues the formal award.
NSF states it aims to inform applicants of a decision within about six months of submission, though the actual timeline varies significantly by program: SBIR/STTR Phase I decisions typically come faster (roughly three to five months), CAREER award decisions typically take five to seven months given the additional review considerations that mechanism carries, and large infrastructure or multi-year center proposals can take nine to twelve months. Once an award decision is reached, additional factors — including NSF’s own budget and appropriations cycle — can affect exactly when funds become available; see CASRAI’s guide on how NSF’s budget and appropriations process affects award timing for that separate, post-decision timing question.
Frequently asked questions
Do I need an NSF ID before I can start writing a proposal?
You need a Research.gov account (and therefore an NSF ID) before you can prepare a proposal in the system, but you don’t need to wait to start drafting the narrative itself — register early since it’s a one-time step, then draft in parallel.
Can I submit an NSF proposal through Grants.gov instead of Research.gov?
For some opportunities, yes — Grants.gov remains an alternate submission path. But Research.gov is NSF’s native, purpose-built system, and most current solicitations expect or require it; check the specific funding opportunity’s submission instructions rather than assuming either path is universally available.
What’s the actual difference between Intellectual Merit and Broader Impacts?
They are NSF’s two official merit-review criteria, applied to every proposal. Intellectual Merit asks how well the proposed activity advances knowledge within or across fields; Broader Impacts asks how the activity benefits society or contributes to specific, desired societal outcomes (workforce development, public engagement, broadening participation, among others). Both must be explicitly addressed in the Project Summary and the Project Description, not just implied.
Can a graduate student or postdoc apply directly to NSF?
For a standard research grant, no — proposals are submitted by an institution on behalf of an eligible PI, and eligibility as PI is typically restricted by career stage or institutional affiliation. NSF does run mechanisms aimed specifically at trainees and early-career researchers (the Graduate Research Fellowship Program applies directly, for example) with their own separate eligibility and application processes distinct from the institutional-PI process this guide describes.
How long does NSF take to make a funding decision?
NSF’s stated goal is to inform applicants within about six months of submission, but this varies by program — some mechanisms (SBIR/STTR Phase I) typically resolve faster, others (CAREER, large infrastructure awards) typically take longer. See Step 7 above.
