Withdrawing a grant application is a distinct administrative action from either appealing an unfavorable review outcome or requesting prior approval for a change to an already-funded award. It means formally pulling a submitted but not-yet-funded application out of the review or funding-decision pipeline, before National Institutes of Health (NIH) or National Science Foundation (NSF) staff make a funding decision on it. This guide covers how withdrawal actually works at each agency, the timing windows that govern it, and the situations where withdrawing is the right call versus situations where it usually isn’t.
What “withdrawing an application” means — and what it doesn’t
Three actions get confused with each other in research-administration conversation, and they are not interchangeable:
- Withdrawing an application (this guide) — pulling a submitted application before a funding decision is made, so it is never reviewed to completion or never funded. Once withdrawn, the application does not proceed further; there is no partial or provisional withdrawal.
- Appealing a review outcome — a formal challenge filed after NIH peer review has already produced a summary statement and the application was not funded, made on narrow procedural grounds (reviewer bias, conflict of interest, lack of reviewer expertise, or factual error in the review). See CASRAI’s guide to writing a grant appeal letter for that separate process — it starts from a completed review, which withdrawal by definition prevents.
- NIH prior approval — a sponsor-approval requirement for specific changes made after an award has already been issued (rebudgeting above the threshold, a no-cost extension, a change of PD/PI, scope changes, and similar). See CASRAI’s NIH Prior Approval dictionary entry. Confusingly, NIH’s eRA Commons system also routes pre-award application-withdrawal requests through a screen labeled “Prior Approval” in its interface — but that is a system/navigation quirk, not a conceptual overlap. Application withdrawal happens before any award exists; prior approval, in the substantive sense the dictionary term covers, only ever applies after one does.
How to withdraw an NIH application
NIH applicants and their institutions withdraw a submitted application through the Withdrawal Request option inside the Prior Approval module of eRA Commons, NIH’s applicant/recipient portal. Based on NIH’s eRA help documentation and NIH Center for Scientific Review guidance:
- Who can act: Either the Principal Investigator (PI) or the institution’s Signing Official (SO) can initiate a withdrawal request in eRA Commons, but only the SO can submit it to NIH — the same institutional-endorsement structure that governs the original application submission applies to pulling it back.
- What’s required: A justification statement explaining the reason for withdrawal, with the option to attach supporting documentation.
- Timing: Withdrawal through the Prior Approval module is generally available for an application that hasn’t yet proceeded to National Advisory Council or Board review. Once an application has moved past that point, or carries certain later-stage statuses, it typically no longer shows as eligible for self-service withdrawal in the module.
- If the module isn’t available: If NIH is already taking programmatic action on the application in a way that blocks the Prior Approval module, NIH’s guidance directs applicants to route the withdrawal request to the Division of Receipt and Referral instead.
- After withdrawal: eRA Commons sends automatic notifications to both the SO and PI once the withdrawal is processed. Withdrawn applications are also flagged distinctly in the system for institutional tracking purposes.
Because the exact eligibility window and status labels are a system-configuration detail that NIH updates from time to time, confirm current specifics against your institution’s eRA Commons access and, where needed, NIH’s own eRA help pages before initiating a withdrawal.
How to withdraw an NSF proposal
NSF proposal submission and withdrawal both run through Research.gov (NSF’s FastLane legacy system was decommissioned for these actions in September 2023, so Research.gov is now the operative system for currently-submitted proposals). Based on NSF’s published Research.gov guidance:
- Who can act: The PI, the institution’s Sponsored Projects Office (SPO) representative, or the Authorized Organizational Representative (AOR) can each initiate a withdrawal request from the proposal’s record in Research.gov. Only the AOR can approve and formally submit the withdrawal to NSF — mirroring the AOR’s role as the only party who can submit a proposal to NSF in the first place.
- How: From the Submitted Proposals list in Research.gov, select the proposal and use the Withdraw Proposal action within the Proposal Actions section of the proposal’s main page.
- Timing: A submitted proposal can be withdrawn at any point before the cognizant NSF Program Officer makes a funding recommendation. Once that recommendation has been made, the proposal is no longer eligible for withdrawal through this route.
- Effective date: The proposal is considered withdrawn from NSF as soon as the AOR approves the withdrawal action in Research.gov — there is no separate NSF-side confirmation step required before it takes effect.
NIH vs. NSF withdrawal, side by side
| NIH | NSF | |
|---|---|---|
| System | eRA Commons, Prior Approval module → Withdrawal Request | Research.gov, Submitted Proposals → Withdraw Proposal |
| Who can initiate | PI or Signing Official (SO) | PI, Sponsored Projects Office (SPO), or Authorized Organizational Representative (AOR) |
| Who can finalize/submit | SO only | AOR only |
| Latest point it’s generally available | Before National Advisory Council/Board review | Before the Program Officer makes a funding recommendation |
| Documentation required | Justification statement (supporting documents optional) | No separate justification document specified in Research.gov’s published process |
Both agencies converge on the same underlying principle even though the mechanics differ: withdrawal is available up until a substantive funding or advisory decision has been reached, and it requires action from someone with institutional signing authority — a PI cannot unilaterally pull an application without institutional sign-off at either agency.
When withdrawing an application actually makes sense
Common, legitimate reasons research administrators and PIs cite for withdrawing a pending application include:
- A material error was found after submission — a wrong budget figure, a missing or incorrect required document (e.g., Other Support, a compliance certification), or a conflict-of-interest disclosure omission serious enough that letting an inaccurate application proceed to review is worse than pulling it and resubmitting correctly at the next opportunity.
- The project no longer needs the funding — the PI secured funding from another source for the same or overlapping work, and NIH/NSF policy on overlapping or duplicate funding requires resolving the redundancy, sometimes by withdrawing one of the competing applications.
- A required change makes the current version obsolete — the PI is leaving the institution, a key collaborator or facility fell through, or the scope changed enough between submission and review that continuing to pursue funding for the original version no longer makes sense.
- A compliance or eligibility problem surfaced late — for example, an IRB or IACUC approval issue, human subjects protection gap, or eligibility question that can’t be resolved in time through NIH’s Just-in-Time process and would otherwise block an award even if the application scored well.
- Institutional or agency direction — the sponsored programs office or the funding agency itself identifies a duplicate submission, an administrative routing error, or another procedural problem and directs the applicant to withdraw and resubmit.
When it usually doesn’t make sense
Withdrawal is not a way to avoid an anticipated unfavorable outcome, and using it that way usually costs more than it saves:
- To dodge a bad score. Letting NIH peer review run its course produces a summary statement with reviewer critiques — the exact input a strong resubmission (or, on narrow procedural grounds, an appeal) depends on. Withdrawing before review forfeits that feedback entirely, leaving the applicant with no documented basis for what to fix next time. See CASRAI’s guide on writing a grant appeal letter for how the summary statement functions in that separate process, and CASRAI’s comparison of NIH vs. NSF resubmission policy for how each agency treats a subsequent, revised submission.
- As a substitute for a minor correction. Small, correctable issues are often better handled through the agency’s existing correction or Just-in-Time mechanisms than through a full withdrawal-and-resubmit cycle, which can cost an entire funding cycle’s worth of time.
- Without institutional sign-off. Because only an SO (NIH) or AOR (NSF) can finalize a withdrawal, a PI-initiated request that skips the sponsored programs office isn’t just against protocol — it typically can’t be completed at all through the standard system route.
Frequently asked questions
Can a PI withdraw an NIH or NSF application without institutional involvement?
No. At NIH, a PI can initiate a withdrawal request in eRA Commons, but only the institution’s Signing Official can submit it. At NSF, a PI can likewise initiate a request in Research.gov, but only the Authorized Organizational Representative can approve and submit it. In practice this means every withdrawal goes through the sponsored programs office regardless of who starts the request.
Does withdrawing an application hurt a future submission?
Neither agency’s published guidance frames withdrawal itself as a penalty or negative mark. It is a routine administrative action. What it does forfeit is any review outcome for that specific submission — there is no summary statement or reviewer feedback to build a resubmission around, since the application was never fully reviewed.
Can a withdrawn application be resubmitted?
Generally yes, as a new submission through the normal application process, subject to the receiving agency’s standard submission and (where applicable) resubmission rules — see CASRAI’s NIH vs. NSF resubmission policy comparison for how the two agencies differ on resubmission specifically.
Is there a deadline within a submission cycle for withdrawing?
The operative limit isn’t a fixed calendar deadline but a process milestone: at NIH, generally before the application proceeds to National Advisory Council or Board review; at NSF, any time before the Program Officer makes a funding recommendation. Once either milestone passes, the application typically is no longer eligible for straightforward self-service withdrawal.
What if NIH is already taking action on the application and the Prior Approval module won’t let me withdraw it?
NIH’s guidance for that situation is to route the withdrawal request directly to NIH’s Division of Receipt and Referral rather than through the self-service module.
This guide describes the general administrative process as published by NIH (eRA Commons / Center for Scientific Review) and NSF (Research.gov). Specific screens, statuses, and eligibility windows are system details that both agencies periodically update — confirm current mechanics against your institution’s eRA Commons or Research.gov access, or your sponsored programs office, before initiating a withdrawal.
