Skip to content

How NSF Grant Review Works: Merit Review Guide

Grantsights12 min readLast updated October 2026

Quick Answer: NSF first checks whether a proposal is complete and compliant. For most compliant proposals, at least 3 external experts review it through ad hoc review, a panel, or both.

In this guide (12 sections)

Get deadline alerts for these grants

Enter your email for a weekly email of upcoming grant deadlines.

No spam. A weekly email with upcoming grant deadlines.

Full answer: NSF first checks whether a proposal is complete and compliant. For most compliant proposals, at least 3 external experts review it through ad hoc review, a panel, or both. They apply Intellectual Merit and Broader Impacts, plus any solicitation-specific criteria. The program officer analyzes the review record and makes an award or decline recommendation to the division director. NSF strives to notify applicants within 6 months, but program and award-processing details can change the timing. This is for faculty, research staff, and new PIs preparing an NSF submission or reading a decision notice.

How NSF Grant Review Works: Current Merit Review and Funding Decisions

Data note: This guide was checked on July 26, 2026 against NSF's current merit-review and proposal-and-award-process pages, PAPPG 24-1 Supplements 1 and 2, and the active PAPPG index (rechecked October 9, 2026). NSF deferred PAPPG 26-1, so PAPPG 24-1 and its supplements remain the current policy layer. NSF has also released draft Guidance on Financial Assistance that will replace the PAPPG (feedback closed August 24, 2026; no effective date yet).

From Manusights, our sister product

Writing an NSF proposal? Manusights offers an AI critique of your proposal against the review criteria in your solicitation: $149 for a research proposal, $49 for a GRFP or other fellowship application. Back within 3 business days of your upload.

Order a grant review

July 26, 2026: the current decision path

StageWhat NSF says happensWhat the applicant should do
Preliminary reviewThe program officer checks completeness and compliance. A noncompliant proposal may be returned without review.Use the solicitation and PAPPG, then confirm every required document before submission.
Merit reviewExternal ad hoc reviewers, a panel, or both usually review compliant proposals.Write to Intellectual Merit and Broader Impacts, plus every solicitation-specific criterion.
Program recommendationThe program officer analyzes review input and relevant portfolio factors, then recommends award or decline.Do not infer an award from a reviewer comment or informal conversation.
Division decisionThe division director generally completes final programmatic approval.Watch Research.gov for an electronic decision.
Award processingNSF's Division of Grants and Agreements completes administrative review and issues an award when applicable.Do not begin costs or commitments until there is an official award.

NSF describes the process as fair, competitive, transparent, and in-depth. The proposal-and-award-process overview is the best current starting point because it distinguishes the scientific review, program recommendation, division approval, and administrative award processing steps.

What controls the review

The funding opportunity controls the program-specific submission and review requirements. The active PAPPG controls the general NSF proposal rules. For the current policy period, NSF has deferred PAPPG 26-1 and published PAPPG 24-1 Supplement 1 (NSF 26-200, awards on or after December 8, 2025) and Supplement 2 (NSF 26-202, awards on or after January 22, 2026). Before drafting, identify the exact solicitation, its extra criteria, and the PAPPG that NSF says applies to that due date.

The hard truth is that a technically strong narrative cannot cure a missing required document or a proposal sent to the wrong program. NSF may return a proposal without review if it does not comply with the solicitation or PAPPG. That is a different outcome from a declined, reviewed proposal, so the first decision is always a compliance and program-fit check.

Applicants often spend their last review round polishing technical prose when the next action should be to compare the proposal package line by line against the solicitation. For example, a PI with a solid research plan but an unconfirmed data-management document should resolve the required-document issue before seeking another scientific edit.

The two merit review criteria

Every NSF proposal is evaluated against two National Science Board-approved criteria.

Intellectual Merit

Intellectual Merit asks about the potential for the proposed project to advance knowledge and understanding within its field or across fields. NSF asks reviewers to consider whether proposed activities explore creative, original, or potentially field-changing concepts; whether the plan is well-reasoned and organized; whether it has a mechanism to assess success; whether the people involved are qualified; and whether adequate resources are available.

Broader Impacts

Broader Impacts asks about the potential for the project to benefit society and contribute to desired societal outcomes. The same five elements apply. A proposal should explain the activity, intended outcome, assessment approach, people responsible, resources, and relationship to the research. Do not treat a generic promise of public benefit as a substitute for an executable plan.

Some solicitations contain additional review criteria. They are not optional. Make a checklist that pairs each stated criterion with the part of the proposal that answers it.

Step 1: preliminary review and reviewer selection

Once NSF receives a proposal, a program officer performs a preliminary review for completeness and conformity with the PAPPG and any applicable solicitation. If requirements are not met, NSF may return the proposal without review.

For most compliant proposals, the program officer identifies external reviewers with relevant subject-matter expertise. NSF's overview says there are at least three external reviewers in most cases. The review can use ad hoc reviewers, a panel of experts, or both. Site visits are used for some programs. RAPID, EAGER, and RAISE proposals are examples that NSF says are internally reviewed only; some travel proposals may also be internally reviewed.

Reviewer assignments and the method of review are program-specific. Do not assume a standard number of panelists, a fixed number of reviews, or a set sequence of panel meetings. NSF program officers are responsible for addressing potentially disqualifying reviewer conflicts of interest.

Step 2: external review

External reviewers provide evaluations that help the program officer decide whether to recommend funding. A panel is not a universal requirement, and a panel discussion does not itself make the funding decision. The reviewers apply Intellectual Merit and Broader Impacts, along with any additional criteria in the solicitation. According to NSF, the 2 criteria apply to every proposal, while a solicitation may add criteria of its own.

The practical writing test is simple: a reviewer should be able to locate the evidence for each criterion without reconstructing it from scattered claims. Place the research question, method, assessment approach, qualifications, resources, and societal outcomes where the solicitation and PAPPG expect them.

One common mistake is assuming that a positive review settles the outcome. NSF explicitly says that all Excellent ratings do not guarantee funding, while a proposal need not receive all Excellent ratings to be funded. The review record matters, but it's not a score-only competition. Don't delay a fit check just because an informal reviewer liked the science.

What NSF Doesn't Tell You

Official sources explain the review stages, but they don't select the right corrective action for a specific proposal. Grantsights' review method is to separate problems that can prevent review from problems that affect merit discussion. A missing required document is a compliance problem. An unconvincing assessment plan is a merit problem. Mixing them creates wasted revision time. According to NSF, most compliant proposals receive input from at least 3 external reviewers, but that does not remove the need to solve the compliance question first.

For example, one PI preparing a cross-disciplinary proposal may have strong preliminary results but an unclear program destination. The better route is to confirm fit with the cognizant program officer and update the cover-sheet selection before investing in a mock review. A PI with a clear program fit and a vague broader-impacts plan has the opposite next action: make the activities, partners, intended outcomes, and assessment method visible in the Project Description. You can compare current opportunities in the Grantsights database and use the NSF grants guide for broader program context.

Step 3: program officer recommendation

The program officer reviews the proposal and analyzes the external-review input. In developing a funded portfolio, NSF says officers may consider factors such as different approaches to significant research and education questions, potential for high-risk transformational advances, capacity building in a promising area, special program objectives, current proposals, and previously funded projects.

After the scientific, technical, and programmatic reviews, the program officer makes an award or decline recommendation to the division director. This is why an applicant should not treat labels such as "fund if possible" as an NSF-wide decision category. The official record is the decision released through Research.gov. A fixed recommendation label can't replace the actual notice and review materials.

Step 4: division decision and award processing

Final programmatic approval is generally completed at the NSF division level by the division director. If a recommendation is approved, it moves to the Division of Grants and Agreements for business, financial, and policy review, processing, and issuance.

NSF generally makes awards to academic institutions within 30 days after the program division or office makes its recommendation. New awardees, cooperative agreements, coordination with another funder, and unusual arrangements can need additional review. Only an NSF grants and agreements officer can make a commitment, obligation, or award on NSF's behalf. A technical or budget conversation with a program officer is not an award authorization.

What appears when NSF makes a decision

NSF releases the decision electronically to the PI through Research.gov. The materials include the context in which the proposal was reviewed, copies of reviews used in the decision with reviewer-identifying material redacted, a panel summary when a panel reviewed the proposal, and site-visit reports when applicable. According to NSF, these materials are released when the decision is made, whether it is an award or a declination.

If the proposal is declined, NSF provides the reason or reasons and the reviews considered. The PI can request additional information from the cognizant program officer or division director. Keep the review materials with the original proposal, then create a revision matrix: one row for each substantive concern, one column for the proposed response, and one for the revised proposal section. This turns feedback into a concrete resubmission decision instead of a general rewrite. For writing changes, see how to respond to grant reviewer comments and how grant reviewers score applications.

Reconsideration and resubmission

A PI who is not satisfied that the proposal was fairly handled and reasonably reviewed may request reconsideration under PAPPG Chapter IV.D. That path is for review fairness and process, not simply disagreement with a scientific assessment. Consult the current PAPPG and the cognizant program before acting because the requirements are procedural. It isn't a way to relitigate scientific judgment.

A declined proposal may be resubmitted after substantial revision, subject to the solicitation and PAPPG. If the concern is an unaddressed requirement, a poor program fit, or a weak evidence trail, a lightly edited resubmission is usually the wrong route. Change the proposal enough that it answers the review record and meets the current opportunity's rules.

A decision checklist before and after review

Before submission:

  1. Confirm that the solicitation is open and that the project fits the named program.
  2. Match every solicitation-specific requirement to a proposal section or document.
  3. Check the applicable PAPPG and supplements, not an older template.
  4. Explain both merit criteria and show how success will be assessed.
  5. Confirm institutional routing, Research.gov or Grants.gov access, SAM registration, and UEI status.

The NSF grants guide can help keep these package-level checks distinct from scientific edits. If you want to check program fit first, Grantsights pricing describes Grantsights Pro and the full match plan it opens.

After a decision:

  1. Download the Research.gov notice, reviews, panel summary, and any site-visit material.
  2. Separate process concerns from scientific or programmatic concerns.
  3. Request reconsideration only when the PAPPG standard and evidence support it.
  4. For a resubmission, map each substantive concern to a material revision and recheck the new solicitation.

Sources

G

Published by

Grantsights

Grantsights editorial content is reviewed against current program pages, award lists, notices, regulations, and application guidance before publication.

Source review

Each guide names the primary sources used for its program. Status, deadlines, eligibility, application routes, and award figures are checked against the source that owns that fact before publication. The page identifies its review date and separates current rules from historical planning evidence.

  • Primary sources selected from the agency, notice, regulation, or award system that owns each claim
  • Status, deadline, eligibility, and award figures rechecked at the page's displayed source-review date

Editorial accountability

Erik Jia, Founder and Editorial Lead

Accountable for source standards, corrections, product claims, and data-source transparency.

Source trust

How this guide is reviewed

Each guide is checked against the official notice, agency pages, and federal award records, and it's dated when those facts change.

Frequently Asked Questions

How long does NSF review take from submission to decision?

NSF strives to tell applicants whether a proposal was declined or recommended for funding within six months. Actual timing depends on the program, review method, award processing, and any special administrative review. Check Research.gov for the decision rather than relying on a fixed panel or notification date.

Who reviews NSF proposals?

A program officer first checks compliance. For most compliant proposals, NSF uses at least three external experts through ad hoc review, a panel, or both. Some proposal categories are internally reviewed, and some programs also use site visits.

What is the difference between ad hoc and panel review at NSF?

Ad hoc reviewers provide individual external reviews. A panel brings external experts together to review proposals. NSF may use either method or a combination, depending on the program. If a panel reviewed the proposal, the PI receives a panel summary with the decision materials in Research.gov.

Can I appeal an NSF funding decision?

A PI who believes the proposal was not fairly handled or reasonably reviewed may request reconsideration under PAPPG Chapter IV.D. A declined proposal may be resubmitted after substantial revision, subject to the relevant solicitation and PAPPG rules.

Does an Excellent rating guarantee NSF funding?

No. NSF states that a proposal does not need all Excellent ratings to be funded, and all Excellent ratings do not guarantee funding. Review input is considered with programmatic factors and the agency's funded portfolio.

Last updated: October 9, 2026. This page is reviewed regularly and updated when eligibility requirements, deadlines, or funding amounts change.

Get deadline alerts for grants like these

Enter your email for a weekly email of upcoming grant deadlines. When this guide has a current verified report, the confirmation email links to its free sample. Available reports are $49 each.

No spam. A weekly email with upcoming grant deadlines.