How Many SBIR Proposals You Can Submit, by Agency

The National Science Foundation’s full-proposal deadline for SBIR and STTR falls tomorrow, November 4, 2026. A company that already has a proposal sitting with NSF cannot use it, and the reason has nothing to do with money.

Per the record for NSF 26-510, your company may have only one project under NSF review at a time, may submit two Project Pitches in any 12 months, and may submit two full proposals per fiscal year. Those are the binding constraints on a founder with more than one fundable idea, and they are not in the award table.

Every agency running SBIR and STTR now publishes a limit like that. The thing to know before you start writing is that no two of them count in the same unit.

NSF Counts Projects Under Review, Not Awards

NSF 26-510 is the main America’s Seed Fund solicitation. Per the record, Phase I pays up to $305,000 over 6 to 18 months, including indirect costs, a fee of up to 7%, up to $6,500 of technical and business assistance, and up to $25,000 for optional I-Corps training. Phase II pays up to $1,250,000, typically over 24 months, with a fee of up to 10% and optional assistance of up to $50,000. Fast-Track combines both in one award of up to $1,555,000 — up to $400,000 for a 6-to-12-month Phase I component and up to $1,155,000 for an 18-to-24-month Phase II component. All three show a November 4, 2026 deadline, and NSF runs three full-proposal deadlines a year: the first Wednesday in November, the first Thursday in March, the first Wednesday in July.

The counting rules sit underneath that. One project under NSF review at a time. Two Project Pitches in any 12 months. Two full proposals per fiscal year. Phase I is pitch-gated, so an idea consumes a pitch before it ever consumes a proposal, and per the record NSF answers a pitch in about one to two months, with an invitation valid for the next two full-proposal deadlines.

At Phase II the unit changes again: per the record, NSF allows one Phase II proposal per Phase I award, and a declined Phase II cannot be resubmitted. There is no second attempt to plan for. Fast-Track has its own version of the same hard edge — per the record, the Phase II money is not automatic, and at the end of the Phase I component you must pass a progress review and a cost review or the award ends there.

What this means in practice: at NSF the scarce resource is a review slot, and a submission spends it whether or not the submission was ready.

DOE Counts Pitches, and a Declined Pitch Is Spent

Since April 2026, per the record, the Department of Energy’s SBIR and STTR programs are run by the Office of Technology Commercialization, with applications handled on the DOE SBIR/STTR Application Hub by its partnership intermediary, ConnectWerx. New companies enter only through a Phase I call that starts with a short pitch. Invited pitches then file a full application for up to $250,000 over 6 to 12 months, or $256,500 with the $6,500 assistance request.

The limit is on pitches: per the record, up to three pitches per company, four open-text questions each with strict word counts. The FY26 Genesis Mission call — four AI-for-science topics, roughly 40 awards from $10 million — capped pitches at three per company and closed September 10, 2026. DOE says a broader Phase I call follows in fall 2026; check the hub for the current call rather than assuming the Genesis terms carry over.

Two details make those three pitches worth rationing. Per the record, a declined pitch cannot be resubmitted, and the topic and the prime company cannot change between the pitch and the full application. A pitch is therefore a commitment to a specific idea under a specific entity, and a thin one burns a third of your access for the call.

NASA Counts Proposals per Appendix, and the Count Resets

Since April 17, 2026, per the record, NASA buys SBIR, STTR and SBIR Ignite research through a single Broad Agency Announcement, 80NSSC26R0003, effective through September 30, 2027, and opens Phase I subtopics in two to three appendices a year. Phase I contracts pay up to $225,000 including up to $6,500 of assistance, over 6 months for SBIR and 13 months for STTR, submitted through ProSAMS.

Per the record, each 2026 appendix accepted no more than two proposals per firm, and NASA says limits reset with each new appendix. That makes the calendar, not the fiscal year, the thing that governs your capacity: the first three appendices — 26A SBIR, 26B SBIR and 26B STTR — ran April 21 to May 21, 2026 and produced 204 selections, and per the record the 2026 windows were open only 30 days each.

A 30-day window with a two-proposal ceiling and a 15-page limit on the technical proposal is a planning problem more than a writing problem. The firms that use both slots are the ones that had subtopic candidates identified before the appendix posted.

The Army Counts per Topic, and a Second Submission Replaces the First

The Army SBIR|STTR program — Army FUZE — releases Focused and Open Topics through the monthly FY2027 Department of War SBIR and STTR Commercial Solutions Openings on DSIP. Per the record, Phase I contracts pay up to $300,000 for up to six months, Phase I awardees may then propose a Phase II of up to $2 million over 12 to 18 months under the same topic, and each topic states its own ceiling.

Here the unit is the topic, and the rule has a trap in it. Per the record, from FY2027 you may send one proposal per topic, and if you submit twice only the latest is evaluated. The common habit of submitting a safe version early and a better version later does not produce two reviewed proposals at DSIP; it produces one, and it is whichever you filed last.

Two further ceilings in the same record are worth knowing before you build a strategy around volume. Firms that miss the Small Business Administration’s performance benchmarks are capped at 20 Army Phase I and Direct to Phase II awards. And companies may hold only two Phase II awards per topic across all agencies — a limit on holdings rather than submissions, which is a different kind of constraint and one that can bind years after the proposal that created it.

At the Air Force the Limit Is the Topic List

AFWERX runs SBIR and STTR for the Department of the Air Force, including SpaceWERX for Space Force topics. Per the record, Phase I awards reach up to $323,000 for up to six months and Phase II up to $2.1 million for up to 24 months, and FY27 Release 1 — SBIR topics 27.BZ and 27.BX, STTR topic 27.TZ — opened October 28 and closes November 25, 2026.

The constraint here is not a count of proposals but what you are permitted to answer. Per the record, you must answer a published topic rather than propose your own idea. In SBIR Release 1 that means 13 Air Force topics, of which 6 accept Phase I proposals and 7 are Direct to Phase II only. The STTR release carries 3 topics, all Direct to Phase II, and Direct to Phase II requires a Department of the Air Force customer and a signed Customer Memorandum. SpaceWERX runs its own bands: per the record, Phase I of $75,000 to $180,000 over 3 to 12 months, and Phase II or Direct to Phase II of $1.25 million to $1.8 million over up to 24 months, with open-topic Phase II needing a signed Space Force customer memorandum.

So the question to ask about an Air Force release is not how many proposals you may file. It is how many of the published topics you can credibly answer, and whether you have the customer signature that half of them require.

Above Phase II, the Gate Is a Person

The largest awards in the current solicitations are not things you apply to in the ordinary sense.

Per the record, the NSF SBIR/STTR Strategic Breakthrough award under NSF 26-510 reaches up to $30,000,000 per company over a project of up to 48 months, and it is invite-only: Phase II awardees submit in Research.gov only on the recommendation of their cognizant Program Officer. The record states that the authority comes from the Small Business Innovation and Economic Security Act (P.L. 119-83), which requires at least one prior Phase II award, matching funds of at least 100% from new private capital or new non-SBIR government funding, and market research showing the technology is an effective solution. Two details in that record matter for planning: the law counts only new money that comes as a result of the award, not funds you already hold, and the authority ends September 30, 2031 unless Congress extends it.

The same shape appears elsewhere. AFWERX offers TACFI of $375,000 to $2 million, STRATFI of $3 million to $15 million, and Strategic Breakthrough awards of up to $30 million to Phase II companies, all with matching requirements, per the record. The NIH Phase IIB Strategic Breakthrough award under the Parent R44 bridges the gap between the end of an SBIR or STTR Phase II and commercialization; the record carries a deadline of April 5, 2029 and no award figure, so see the listing for amounts.

None of these consumes a submission slot. They consume a relationship with a program officer and a committed co-funder, which is a longer lead time than any deadline on this page.

The Check Before You Write

Five minutes, in this order, for any SBIR or STTR opportunity:

1. Find the limit and its unit. Projects under review, pitches, proposals per appendix, proposals per topic, awards held. The number is meaningless without the unit.
2. Count what is already in flight at that agency. At NSF a single project under review closes the door until it clears. Your pipeline, not the solicitation, decides whether you are eligible to submit this week.
3. Ask whether a decline is terminal. A declined NSF Phase II cannot be resubmitted. A declined DOE pitch cannot be resubmitted. Where that is true, the first attempt is the only attempt.
4. Find the reset. NASA’s limits reset with each appendix. NSF’s proposal count runs by fiscal year, its pitch count by rolling 12 months. Those are different clocks and they expire differently.
5. Confirm the registrations are live. Per the records above, submission systems differ by agency — Research.gov at NSF, the Application Hub at DOE, ProSAMS at NASA, DSIP at the Army and Air Force — and most sit behind an active SAM.gov unique entity identifier and an SBA Company Registry record. A lapsed registration spends a slot as effectively as a weak proposal.

Questions Founders Ask

Does a Project Pitch count against my proposal limit at NSF? They are separate counts in the same record: two Project Pitches in any 12 months, two full proposals per fiscal year, and one project under NSF review at a time. An invitation is valid for the next two full-proposal deadlines, so a pitch can be spent without a proposal following it.

If my proposal is declined, can I fix it and resubmit? It depends on the agency and the phase, and the open records disagree. NSF allows one Phase II proposal per Phase I award with no resubmission of a decline. DOE states that a declined pitch cannot be resubmitted. Other calls are silent, which is not the same as permission — read the solicitation you are answering.

Can I submit the same work to two agencies? The records above do not set a cross-agency submission limit, but the Army record does cap holdings: companies may hold only two Phase II awards per topic across all agencies. Agencies also require disclosure of current and pending support, so overlapping submissions are visible to reviewers.

Why does one solicitation show an award ceiling and another show nothing? Because the listing reflects what the funder published. Where a record carries no figure, we print “see listing” and link it rather than borrowing a number from a similar program.

The Bottom Line

Reporting on the 2026 SBIR and STTR reauthorization has mostly been about ceilings — the new $30 million tier, the Phase I and Phase II bands. The change that touches more companies is quieter and sits in the FY2027 solicitations now open: each agency limits how many times one firm may come to the window, and each one counts something different. NSF counts projects under review. DOE counts pitches. NASA counts proposals per appendix. The Army counts proposals per topic, and a resubmission overwrites rather than adds.

Read for award size, those nine records look like a menu. Read for the counting rule, they look like a budget — and the scarce item is not the proposal you write but the slot it occupies.

Open SBIR and STTR records across all of these agencies, with their deadlines and terms, sit in the OpenGrants federal grants hub and the SBIR and STTR hub; for companies earlier in the capital stack, the startup grants hub is the better starting point. Subscribe to Funding Friday, our weekly grant digest, for open programs with every figure checked against the listing.

Every figure above was taken from the program record as held in the OpenGrants database on November 3, 2026, and attributed to the record that carries it. Submission limits, award ceilings, deadlines and eligibility language change, and several of the solicitations described here run on monthly or per-appendix cycles; confirm each against the funder’s own listing before you rely on it. OpenGrants does not guarantee funding outcomes; results vary by organization and opportunity.