How to Apply for SBIR Grants and STTR Awards in 2026

To apply for an SBIR grant, match your technology to an agency's open solicitation, read it in full, confirm eligibility, then submit a Phase I proposal before the close date (SBIR.gov: Apply). As of April 2026, agencies may issue Phase I awards up to $323,090 and Phase II awards up to $2,153,927 without SBA approval (SBIR.gov: About SBIR and STTR).

• How to apply for an SBIR grant: the 9-step application process
• Pick the agency by what you are building, not by the nearest deadline
• How much is an SBIR award in 2026: caps versus what agencies offer
• SBIR Phase 1 funding: what a Phase I award actually buys
• STTR vs SBIR: the research partner is the whole difference
• The SBIR timeline: a year-long process you start before you need it
• Non-dilutive funding for US startups: the equity and IP fine print
• Where government grants for startups get lost: the Phase I work plan
• When this matters for your raise
• FAQ

SBIR and STTR are the largest pool of non-dilutive money open to a US deep tech company, and the government takes none of your cap table for it. America's Seed Fund reports approximately 4,000 companies funded per year, $4 billion invested each year, and 0% equity or IP ownership taken (America's Seed Fund).

The phase ladder is the part founders misread. Phase I is not a small cheque on the way to a big one, it is a bounded feasibility test that buys you the right to ask for Phase II. Only Phase I awardees are eligible for Phase II (SBIR.gov FAQ), so the first application is a gate, not a warm-up.

Two decisions carry most of the outcome, and neither is a deadline. The first is which agency you write to, because agencies fund their own missions and their solicitations, not a generic pool of innovation money. The second is when you start, because this is a year-long procurement cycle, not bridge financing.

How to apply for an SBIR grant: the 9-step application process

The SBIR application process is linear, and every step before writing is a filter you can fail cheaply.

1. Match your technology to an agency mission. Pick the agency whose mission funds what you build: NSF for general deep science, DOE for energy and computing, NASA for space, NIH for health, ED/IES for education, DoD components for defense and dual-use.
2. Read the whole solicitation. The official sequence is find the agency solicitation, read it completely, confirm eligibility, prepare the proposal, submit before the closing date, undergo agency review, then receive an award decision (SBIR.gov: Apply).
3. Confirm eligibility and ownership. SBIR is equity-free federal funding for American small businesses (SBIR.gov: About SBIR and STTR), and agencies add their own rules: NSF restricts majority ownership by multiple VC, private equity or hedge funds (NSF: SBIR/STTR Eligibility).
4. Register in the federal systems the solicitation names. SAM.gov, SBIR.gov and the agency portal. Do this the day you pick the solicitation, not the week of the deadline.
5. Submit the pre-proposal step if the agency has one. NSF requires a Project Pitch before a full proposal, accepts it at any time, and returns feedback in about one to two months (NSF: How It Works).
6. Decide SBIR or STTR before writing the work plan. SBIR permits a research partner; STTR requires a nonprofit research institution, a defined work split and a co-PI at the partner (NIH Seed: Understanding SBIR and STTR).
7. Scope the budget to the solicitation, not the statutory cap. NSF's FY2026 Phase I offers up to $305,000 (NSF Solicitation 26-510), while DOE's FY26 Genesis Mission limits full applications to $250,000, or $256,500 with TABA (DOE/ConnectWerx FY26 Genesis Mission).
8. Write the aim as a falsifiable technical hypothesis. NASA describes Phase I as the jumping-off point for establishing scientific, technical and commercial merit and feasibility (NASA: SBIR/STTR Phase I). Name the measurement and the threshold.
9. Submit before close, then start Phase II early. At NSF, a Phase II proposal may go to an open opportunity 6 to 24 months after the Phase I starts, and the Phase I does not need to be finished first (NSF: SBIR/STTR Eligibility).

Do not run this in parallel across four agencies on your first attempt. Each solicitation carries its own topics, forms, budget rules and review criteria, and a proposal rewritten to fit a second agency at the last minute reads exactly like what it is.

Pick the agency by what you are building, not by the nearest deadline

Agency choice is a technical-fit decision, and getting it wrong wastes the whole quarter regardless of how well you write.

Read the topic list before the eligibility page. Most agencies fund against published topics, and the fastest way to know whether you have a real shot is whether a topic describes your work in the agency's own words.

Do not pick the agency with the closest deadline. A near deadline is the worst reason to choose a funder whose mission does not cover your technology, and the review panel is made of people who work on that mission every day.

How much is an SBIR award in 2026: caps versus what agencies offer

The cap is not the offer, and confusing the two is the most common budgeting error in a first application.

The ceilings are statutory headroom, not a target. Awards above the April 2026 Phase I and Phase II caps require an SBA waiver (SBIR.gov: About SBIR and STTR), which tells you those numbers describe the outer edge of the program rather than what a given panel will fund.

Build the budget from the work, then check it against the solicitation limit. A budget that lands suspiciously close to the exact request limit, with round numbers in every line, invites the reviewer to audit your assumptions instead of your science.

SBIR Phase 1 funding: what a Phase I award actually buys

SBIR Phase 1 funding buys a bounded feasibility test and evidence for a Phase II proposal. It does not buy runway.

NASA describes Phase I as the jumping-off point for establishing an innovation's scientific, technical and commercial merit and feasibility (NASA: SBIR/STTR Phase I). Read that literally: the deliverable is a finding, and a negative finding delivered on time is a completed Phase I.

• What it covers: the specific experiment, prototype or study described in your work plan, on the timeline the solicitation sets, at the amount that solicitation offers.
• What it does not cover: the general operating cost of a startup that happens to be doing research. The award is scoped to the project, not to the company.
• What it sets up: eligibility for Phase II, which is the real money in the ladder and is only open to Phase I awardees (SBIR.gov FAQ).

Phase I is not bridge money. It is a year-long procurement process that pays you to test a hypothesis, and the payoff is the option on Phase II.

Do not build a runway plan that depends on a Phase I decision. Agencies review on their own calendar, the amounts are set by the solicitation rather than by your burn, and a feasibility award cannot be stretched into a general-purpose seed round.

STTR vs SBIR: the research partner is the whole difference

In the sttr vs sbir choice, one question decides it: does a nonprofit research institution have to do part of the work?

Sources: NIH Seed: Understanding SBIR and STTR and NSF Solicitation 26-510.

Pick STTR when the science genuinely lives in a lab you cannot move. The instrument, the cell line, the beam time or the professor is the reason to accept a mandated work split and a second PI.

Do not add a university partner to look credible on an SBIR application. STTR's work-share floors are structural commitments, and once you take them on, a percentage of the award and a percentage of the schedule sit outside your control.

The SBIR timeline: a year-long process you start before you need it

Time to cash is the number founders skip, and it is measured in quarters.

NSF publishes the clearest version: Project Pitch feedback in about one to two months, then notification about funding roughly six months after the proposal goes in (NSF: How It Works). The Phase I performance period then runs 6 to 18 months under NSF's FY2026 opportunity (NSF Solicitation 26-510), and 6 to 12 months in SBIR.gov's general description (SBIR.gov: Apply).

Project Pitch submitted            month 0
Pitch feedback                     month 1-2       (NSF)
Full proposal submitted            month 3-4
Funding notification               ~6 months after submission (NSF)
Phase I performance period         6-18 months     (NSF FY2026)
Phase II proposal window opens     6-24 months after Phase I start

Start the Phase II proposal while Phase I is still running. At NSF, a Phase II proposal may be submitted 6 to 24 months after the preceding Phase I starts, and the Phase I does not need to be complete (NSF: SBIR/STTR Eligibility). Founders who wait for the final report lose a cycle for no reason.

Do not treat the phase duration as time to cash. Review times vary by agency, and the gap between submitting and knowing is long enough that a company running on three months of runway should be raising, not applying.

The cap is not the offer, and the award is not the runway. SBIR costs you 0% of your equity and several quarters of your calendar.

Non-dilutive funding for US startups: the equity and IP fine print

Non-dilutive funding for US startups is real here: the government takes 0% equity and 0% IP ownership (America's Seed Fund). The conditions sit elsewhere.

• Data rights stay with you: the small-business awardee owns and has full right and title to data developed under an SBIR award, and any additional government data rights have to be negotiated after award rather than made a condition of receiving it (SBIR.gov FAQ).
• Ownership rules can disqualify you: NSF restricts majority ownership by multiple VC, private equity or hedge funds (NSF: SBIR/STTR Eligibility), so a cap table built for one purpose can close the door on the other.
• The waiver line is a signal: awards above the April 2026 caps require SBA approval (SBIR.gov: About SBIR and STTR), which is a reminder that this is federal money with federal process attached.

Check the ownership rules before you sign the term sheet, not after. If institutional investors will hold a majority after your next round, the eligibility question is a live one and it belongs in the diligence conversation.

Do not let anyone tell you the award costs nothing. It costs a quarter of founder time on the application, reporting obligations for the life of the award, and a work plan you are committed to executing.

Where government grants for startups get lost: the Phase I work plan

Government grants for startups are won on the work plan, and lost on aims that read like a funding request.

Reviewers are assessing scientific, technical and commercial merit and feasibility (NASA: SBIR/STTR Phase I). An aim they cannot judge as true or false gives them nothing to score.

✅ Good: "Phase I tests whether the catalyst holds selectivity above the threshold after 500 hours at operating temperature. If it does not, the program stops and we report the negative result." It states a measurement, a threshold and a stop condition.

❌ Bad: "Phase I funds our engineering team while we build the product and get ready to raise a seed round." It describes a cashflow benefit, and it tells the panel the award is a substitute for a runway plan.

Three failure modes account for most weak first applications:

• Aims that cannot fail: if every possible result counts as success, there is no feasibility question and no reason to fund the study.
• Budgets written to the ceiling: costs that scale to the request limit rather than to the work invite scrutiny of every line.
• Ignoring the agency's own words: the topic description is the scoring rubric in prose, and a proposal that never uses the topic's vocabulary reads as off-mission.

Write the negative-result plan into the proposal. Saying what you will conclude if the experiment fails is the single clearest signal that you understand a feasibility award, and most first-time applicants leave it out.

When this matters for your raise

Grant money changes the shape of a round without removing the need for one. A Phase I award pays for technical de-risking you would otherwise fund out of the seed, at 0% equity (America's Seed Fund), and the milestone list you write for reviewers is close to the one an investor will interrogate.

Timing is the part to get right: the notification wait means SBIR complements a raise rather than replacing it, and ownership rules can cut the other way if institutional investors take majority control (NSF: SBIR/STTR Eligibility). When you do go out, Causo matches that milestone set against investors actively deploying in your sector and drafts the outreach. If you are weighing other non-dilutive routes, read the Activate Fellowship application guide for hard tech fellowships, the AI2 Incubator application guide for AI-specific programs, or venture debt at seed for the debt side of the same question.

FAQ

How much is an SBIR Phase I award? It depends on the agency, not on one national number. As of April 2026, an agency may issue a Phase I award, including modifications, up to $323,090 without SBA approval (SBIR.gov: About SBIR and STTR), while SBIR.gov's general guide describes Phase I as 6 to 12 months at $50,000 to $275,000 (SBIR.gov: Apply). Live solicitations sit below that ceiling: NSF's FY2026 standard Phase I offers up to $305,000 (NSF Solicitation 26-510) and DOE's FY26 Genesis Mission limits full applications to $250,000, or $256,500 with TABA (DOE/ConnectWerx FY26 Genesis Mission). Take your number from the solicitation you are applying to.

What is the difference between SBIR and STTR? The research partner. SBIR allows a partnership but does not require one, and lets the small business outsource up to 33% of Phase I research and 50% of Phase II research; STTR requires a nonprofit research-institution partner, with at least 40% of the work performed by the small business and 30% by the partner (NIH Seed: Understanding SBIR and STTR). STTR also changes who runs the project: at NSF it requires a partner institution, a subaward, a small-business PI and a co-PI at the partner (NSF Solicitation 26-510). Choose STTR when the core science lives in a lab you cannot move in-house.

How long does SBIR funding take? Months, and the wait varies by agency. At NSF, Project Pitch feedback arrives in about one to two months, and applicants are notified about funding roughly six months after submitting the proposal (NSF: How It Works). The performance period then runs 6 to 12 months in SBIR.gov's general description of Phase I (SBIR.gov: Apply) and 6 to 18 months under NSF's FY2026 opportunity (NSF Solicitation 26-510). Start roughly a year before you need the cash to land.

Does SBIR take equity? No. America's Seed Fund reports approximately 4,000 companies funded per year, $4 billion invested each year, and 0% equity or IP ownership taken by the government (America's Seed Fund). The awardee owns and has full right and title to data developed under an SBIR award, and any additional government data rights must be negotiated after award rather than made a condition of receiving it (SBIR.gov FAQ). Non-dilutive is not unconditional: NSF restricts majority ownership by multiple VC, private equity or hedge funds, so check the eligibility page before you assume you qualify (NSF: SBIR/STTR Eligibility).

What are the three phases of the SBIR Program? Phase I tests feasibility, Phase II builds the technology, Phase III is commercialization. SBIR.gov's general guide describes Phase I as 6 to 12 months at $50,000 to $275,000 and Phase II as 24 months at $400,000 to $1.8 million (SBIR.gov: Apply), and only Phase I awardees are eligible for Phase II (SBIR.gov FAQ). Phase III is where you sell the result, through procurement or the private market, rather than a cheque you apply for. Some agencies also publish a direct route into Phase II: ED/IES offered a $1,000,000 Direct-to-Phase-II option in its FY2026 cycle (ED/IES FY2026 SBIR Solicitation).