NSF SBIR Guide 2026 | Win America's Seed Fund

Full answer: NSF SBIR (branded as "America's Seed Fund") provides up to $305,000 for Phase I and up to $1,250,000 for Phase II to small businesses developing deep technology innovations, plus a Fast-Track option that combines both phases. NSF's SBIR program differs from NIH and DOD versions by focusing on engineering, physical sciences, and computing rather than health or defense. Full proposals require an invitation from a Project Pitch first. Applications are reviewed by technical panels, and NSF places heavy emphasis on both technical innovation and commercial potential.

NSF SBIR Guide: America's Seed Fund Phase I, Phase II, and How to Win (2026)

Data note: Award amounts come from NSF solicitation NSF 26-510. Award counts come from USASpending.gov FY2025 data. Program parameters may change with new solicitations.

Source review by Grantsights, June 3, 2026: We rechecked NSF America's Seed Fund pages, Project Pitch instructions, active awardee data, SBIR.gov, NSF funding updates, and the broader NSF grants context. Where official sources stop: NSF now frames Phase I around a required Project Pitch invitation and has raised common maximums to $305,000 for Phase I and $1,250,000 for Phase II, while Phase II still requires Phase I lineage.

Who this is for: Startup founders, small business owners, and R&D leads building technology products in engineering, computing, materials science, advanced manufacturing, or other NSF-supported technical areas. If you're developing something genuinely new (not just an app or a service business) and you're a U.S. small business, the NSF SBIR program is one of the best non-dilutive funding sources available.

NSF's SBIR program is one component of the broader NSF grants portfolio, and it's different from every other agency's version. It's the only federal SBIR that brands itself as a seed fund and acts like one. NSF program directors actively help awardees find customers, connect with investors, and refine their commercial strategy. That's not marketing language. It's how the program actually operates, and it's why many founders who've worked with multiple agencies say NSF's SBIR is the best run.

But NSF is also pickier about what it funds. The technology has to be genuinely novel, not an incremental improvement to an existing product. And the commercialization plan matters more here than at agencies like DOD, where the government itself is often the customer.

This guide covers how NSF SBIR works, how it differs from other agencies, what the review process looks like, and how to build an application that scores well. For the full range of non-dilutive funding options available to small businesses beyond SBIR, see our small business grants complete guide.

Last updated: October 2026

---

NSF SBIR Award Amounts and Structure

NSF structures its SBIR program in two phases, with a Fast-Track option and a Phase IIB supplement. These figures come from NSF 26-510:

Phase I is the entry point. You propose a technical project that demonstrates your innovation is feasible. NSF expects a clear deliverable: proof that your approach works, not a finished product. The up to $305,000 covers salary, equipment, materials, subcontractors, indirect costs, the small business fee, TABA, and optional I-Corps funding for 6 to 18 months of work.

Phase II is the big award. At up to $1,250,000, typically over 24 months, it funds full development: prototyping, testing, iteration, and early commercialization activities. Only companies with a Phase I award can apply for Phase II on the same topic. This is NSF's "lineage" requirement, and it's stricter than some other agencies. The exception is Fast-Track, where a single invited proposal covers a Phase I component (up to $400,000) and a Phase II component (up to $1,155,000).

Phase IIB supplemental funding is where it gets interesting. NSF will match third-party investment (from VCs, angels, strategic partners, or customers) with a supplement of $50,000 to $500,000. If you raise $500,000 in qualifying investment during Phase II, NSF can add up to another $500,000 to your award. Combined with a $1,250,000 Phase II, that can bring your Phase II total to $1.75 million.

---

How NSF SBIR Differs from NIH, DOD, and Other Agencies

Every federal agency with over $100 million in FY2025 in extramural R&D must run an SBIR program. That means 11 agencies offer SBIR funding. But they're not interchangeable. Here's how NSF compares:

The biggest difference is customer. NSF expects your product to sell in commercial markets. NIH assumes your end users are in healthcare. DOD often assumes the government itself will buy what you build. This shapes everything about how you write your proposal and how reviewers evaluate it.

NSF doesn't allow Direct-to-Phase II. At NIH and some DOD components, you can skip Phase I if you've already demonstrated feasibility with your own funds or other grants. NSF still requires a Phase I before Phase II, but its Fast-Track award lets you propose both phases in one proposal.

NSF's commercialization emphasis is the strongest. While every agency says they care about commercial outcomes, NSF means it. Your Phase I proposal needs a commercialization plan, and Phase II reviewers weight commercial potential heavily. Companies that can't articulate who will buy their product, at what price, and through what channel struggle in NSF's review.

---

NSF SBIR Technology Focus Areas

NSF organizes its SBIR topics into technology areas that shift periodically. As of 2026, the main areas include:

• Artificial Intelligence and Machine Learning (AI/ML)
• Advanced Manufacturing and Nanotechnology
• Advanced Materials
• Biomedical Technologies (non-clinical)
• Chemical Technologies
• Communication Systems and Networking
• Cybersecurity
• Education and Training Technologies
• Electronic Hardware, Robotics, and Wireless
• Energy and Sustainability
• Internet of Things (IoT) and Cyber-Physical Systems
• Quantum Technologies
• Semiconductors and Microelectronics
• Space and Geospatial Technologies

The new Technology, Innovation, and Partnerships (TIP) directorate at NSF has expanded some of these areas, particularly in semiconductors, critical minerals, and biotechnology. Check the current NSF SBIR solicitation for the latest topic list, because areas get added and refined each fiscal year.

Your innovation doesn't have to fit neatly into one area. NSF allows you to propose cross-cutting technologies. If your AI-powered materials discovery platform spans AI/ML and Advanced Materials, you pick the area that best matches your core innovation and note the cross-cutting relevance.

---

The NSF SBIR Application: What You Need

Phase I Proposal Components

A Phase I proposal includes:

1. Project Description. Check the current solicitation for the page limit. This is the core of your application. It covers the technical innovation, R&D plan, team qualifications, and commercialization strategy.
2. Commercialization Plan (included in Project Description). NSF wants to see your market analysis, competitive positioning, revenue model, IP strategy, and go-to-market approach. Give it a substantial share of your page limit.
3. Budget and Budget Justification. Detailed line items for how you'll spend up to $305,000. NSF reviewers scrutinize budgets, especially for excessive PI salary percentages and unjustified subcontractor costs.
4. Biographical Sketches. For the PI and all senior personnel.
5. Current and Pending Support. All other funding your team receives or has applied for.
6. Data Management Plan. Brief description of how you'll handle research data.
7. Letters of Support. From potential customers, partners, or technical advisors. These aren't scored directly but strongly influence reviewer perceptions of commercial viability.

What Reviewers Actually Score

NSF SBIR reviewers evaluate proposals on these criteria:

Intellectual Merit (Technical Innovation): Is this genuinely new? Does the R&D plan make sense? Can this team execute it? Reviewers are technical experts in your area, and they can tell the difference between a real advance and a repackaged existing technology.

Broader Impacts (Commercial Potential + Societal Benefit): Will this become a product? Is there a real market? Does the team understand their customers? Will it create jobs, improve infrastructure, or benefit society in measurable ways?

NSF SBIR review panels typically include 3 to 4 reviewers per proposal. They read independently, write preliminary scores and comments, then discuss proposals as a panel. The discussion matters: a lukewarm preliminary review can be upgraded (or a strong one downgraded) based on panel conversation. Program directors observe the panel and use the results to make funding decisions, occasionally overriding panel rankings based on portfolio balance or strategic priorities.

---

America's Seed Fund: More Than Branding

NSF rebranded its SBIR/STTR program as "America's Seed Fund" in 2017, and it wasn't just a name change. The rebrand reflected a strategic shift toward treating SBIR more like venture investing:

I-Corps integration. NSF encourages (and sometimes requires) SBIR applicants to complete the I-Corps customer discovery program. I-Corps teaches you to validate your market assumptions through 100+ customer interviews in seven weeks. Many successful NSF SBIR applicants complete I-Corps before submitting Phase I, and NSF views I-Corps participation favorably during review.

Beat the Odds Bootcamp. NSF runs the "Beat the Odds" bootcamp for prospective SBIR applicants, a multi-day workshop that walks you through the application process, review criteria, and common mistakes. It's free and open to anyone considering an NSF SBIR submission. Attending doesn't guarantee funding. Check the NSF SBIR events page for upcoming sessions.

Demo Days. NSF hosts events where SBIR awardees present to investors, corporate partners, and government buyers. These aren't formalities: companies have closed funding rounds and signed partnership agreements at NSF Demo Days.

Active program management. Unlike some agencies where your program officer is a distant bureaucrat, NSF SBIR program directors often act more like portfolio managers. They'll connect you with other awardees, suggest potential customers, and push you to think about commercialization. This hands-on approach is unusual for government and genuinely useful.

---

NSF SBIR Success Rates and What Drives Them

NSF's current solicitation doesn't publish a Phase I success rate, and odds vary by technology area and submission cycle. You can see who has won in your area through NSF Award Search. A few patterns follow from how NSF reviews proposals:

First-time applicants face a steeper learning curve. Companies submitting their first SBIR proposal of any kind are learning the format, the review criteria, and the commercialization expectations at the same time. Applicants who have written SBIRs at other agencies, or who resubmit after reading reviewer feedback, usually turn in tighter proposals.

Teams with prior NSF connections do better. If your PI or co-PI has worked with NSF before (through research grants, I-Corps, or other programs), reviewers implicitly trust the feasibility arguments more. This isn't bias; it's because experienced PIs write better proposals and understand what NSF values.

Proposals with customer evidence win more often. A letter of intent from a potential customer saying "we would pay $X for this product if it works" is worth more than 10 pages of market analysis. Reviewers are scientists and engineers, not market researchers. Concrete evidence of demand is more persuasive than TAM/SAM/SOM slides.

Resubmissions improve. If you submit, get rejected, incorporate reviewer feedback, and resubmit, your odds go up substantially. NSF provides written reviewer comments for all declined proposals. Read them carefully: they tell you exactly what didn't work.

---

NSF SBIR Timeline: From Idea to Award

NSF accepts Project Pitches at any time, and invited full proposals are due on set deadlines. Under NSF 26-510, the next full-proposal deadline is November 4, 2026, followed by March 4, 2027 and July 7, 2027. The typical timeline:

Total time from first idea to Phase II funding can exceed three years. This is the reality that most startup founders underestimate. SBIR is patient capital, not fast capital. If you need money in three months, this isn't the program for you. If you can plan 12 to 18 months ahead, SBIR is one of the best deals in startup funding: non-dilutive, no strings on IP, and the government's stamp of approval on your technology.

---

The Project Pitch Hurdle That Surprises First-Time NSF SBIR Applicants

The step that most frequently blindsides first-time NSF applicants isn't the full proposal. It's the Project Pitch requirement that gates full proposal submission. Unlike NIH or most DOD SBIR components, NSF requires you to submit a short Project Pitch (roughly 2 pages) and receive an invitation from a Program Director before you can submit a full proposal. Program Directors review pitches and decide whether the technology is within NSF's scope and sufficiently novel. Not every pitch gets an invitation, and each company can submit at most two Project Pitches in a 12-month period. Founders can invest months in a prototype and a full commercialization plan before learning their pitch was declined because their technology was incremental rather than transformational, NSF's language for "we've seen this before." The pitch isn't a formality. It's the first real gate, and it happens before you write a single word of the main proposal.

SBIR/STTR authority lapsed on October 1, 2025, and NSF paused new Project Pitches during the lapse. Congress reauthorized the programs in April 2026, and NSF is accepting pitches again: its Get Started page says you can submit a Project Pitch anytime and hear back in about 1 to 2 months. The current solicitation, NSF 26-510, had a full-proposal deadline of July 27, 2026, which has passed. The next one is November 4, 2026. Because invitations take time, a pitch submitted late in the cycle may only be ready for the March 4, 2027 deadline.

A common reason NSF proposals fail isn't technical weakness. It's the commercialization section. NSF's review panels include both technical experts and commercialization reviewers. The commercialization reviewer isn't impressed by TAM calculations. They're looking for evidence that real customers will pay real money. Signed letters of intent from customers, each specifying a per-unit price they'd pay for a working product, carry more weight than a market size slide.

NSF SBIR Phase I awards go up to $305,000 under the current solicitation, up from $275,000 in earlier solicitations. Phase II awards go up to $1,250,000, with Phase IIB supplements of $50,000 to $500,000 matching third-party investment. Phase IIB is easy to overlook: a company that only learns about it late in Phase II can miss the chance to line up qualifying investment. If you win a Phase II, ask your Program Director about Phase IIB in your first quarterly update, not in month 20.

---

What the NSF SBIR Website Doesn't Tell You

Here's what you won't find in the solicitation.

Reviewers read closely. SBIR panelists do deep technical reads. They'll check your citations, evaluate your methodology, and assess whether your timeline is realistic. Sloppy technical writing or inflated claims get caught. Reviewers may check a PI's publication record during panel discussions to verify expertise claims.

The commercialization plan is where most proposals lose points. Technical sections are usually adequate because scientists and engineers know how to describe their work. But the commercialization section is where applicants stumble. "Our TAM is $50 billion in FY2025" means nothing. "We've spoken with 15 water treatment plant operators in the Southwest, and 12 said they'd pay $20,000 annually for a sensor that detects lead contamination in real time" means everything. Specificity wins.

Program directors have more discretion than you'd think. While panels rank proposals, program directors make final decisions and can fund proposals that weren't top-ranked if they fill a portfolio gap. That can mean funding a lower-ranked proposal because the program director wants representation in an underserved technology area. Conversely, a top-ranked proposal can be declined because the technology overlaps with a recently funded award. This isn't common, but it happens.

Phase IIB matching is underused. The Phase IIB supplement, where NSF matches third-party investment dollar-for-dollar up to $500K, is one of the best deals in government funding. Yet many Phase II awardees don't apply for it because they don't realize it exists or don't have the investor relationships to trigger the match. If you're in Phase II, start building investor relationships early specifically to unlock Phase IIB.

---

Common Mistakes in NSF SBIR Proposals

1. Describing a product, not an innovation. NSF funds R&D, not product development. Your proposal should articulate a technical problem that doesn't have a good solution, explain your novel approach, and describe the R&D needed to prove it works. If your "innovation" is applying existing technology to a new market, NSF won't fund it.

2. Ignoring the competition honestly. Every proposal needs a competitive analysis section, and most applicants either claim no competitors exist (reviewers don't believe this) or list competitors without explaining differentiation. Name your competitors. Acknowledge what they do well. Then explain specifically why your approach is technically superior, not just different.

3. Overloading the budget on PI salary. If 60% of your $305,000 budget is PI salary, reviewers wonder whether you're using SBIR as a salary supplement rather than a real R&D program. Keep PI salary reasonable relative to the technical work described. NSF allows reasonable salaries but scrutinizes proposals where most money goes to one person.

4. Weak letters of support. A generic letter saying "we're interested in this technology" is almost worthless. Strong letters say: "We currently spend $X per year on [problem]. If [company] develops a working prototype, we would evaluate it for deployment in our [specific application] and are prepared to commit to a pilot program." Get specific or don't include the letter.

5. Forgetting broader impacts beyond commerce. While NSF SBIR weights commercialization heavily, the broader impacts criterion still includes societal benefit, workforce development, and broadening participation. Proposals that address environmental sustainability, underserved communities, domestic manufacturing, or STEM workforce training score higher on this criterion.

For more on structuring competitive federal proposals, see our guide on how to write a grant proposal.

---

SBIR vs. STTR: Which Program Should You Use?

STTR (Small Business Technology Transfer) is SBIR's sibling program with one main difference: STTR requires a formal partnership with a research institution (university or federal lab). Here's when to use each:

Use SBIR if: Your team can do the R&D in-house and you meet the ownership requirements.

Use STTR if: You need a university or lab partner for specialized equipment, expertise, or IP.

Many companies are eligible for both and should choose based on where the technical work actually happens. If you're licensing university IP and the professor will do substantial R&D, STTR is the natural fit.

---

After the Award: Making the Most of NSF SBIR

Winning a Phase I is step one. Here's how to maximize the opportunity:

Engage your program director early and often. Send quarterly updates even if not required. Ask for introductions to other awardees in complementary technology areas. Program directors appreciate proactive communication and are more likely to advocate for your Phase II if they've been following your progress.

Complete I-Corps during Phase I. If you haven't done I-Corps yet, apply for it during your Phase I performance period. The customer discovery insights will strengthen your Phase II commercialization plan enormously.

Document everything for Phase II. Phase II reviewers want to see what you accomplished in Phase I. Keep detailed technical records, milestone completion reports, and evidence of commercial traction. A Phase II proposal that says "in Phase I, we achieved X, Y, and Z as proposed" with supporting data is vastly stronger than a vague narrative.

Start Phase II planning early. Under NSF 26-510, you can submit a Phase II proposal between 6 and 24 months after your Phase I start date, so the window can open while Phase I is still running. You don't want to start writing from scratch at that point. Begin drafting your Phase II proposal, gathering letters of support, and building your commercial evidence case while Phase I is still active.

---

What USASpending.gov Shows for "NSF SBIR" Awards

A search of FY2025 USASpending.gov award descriptions containing "NSF SBIR" found 6 awards totaling $6.7M. Treat this as a rough keyword sample: two of the awards below exceed NSF's $1,250,000 Phase II ceiling, so the search likely picked up awards from other agencies or programs that mention NSF SBIR.

Top Recipients

Top States

Data source: USASpending.gov FY2025 award description search, accessed 2026-04-04.

Related Guides

• SBIR Commercialization Plan
• SBIR Phase 2 Guide

---

Sources

• NSF SBIR/STTR Program Page (America's Seed Fund)
• NSF 26-510 SBIR/STTR Phase I, Fast-Track, and Phase II Solicitation
• NSF SBIR Solicitation
• NSF Award Search
• NSF I-Corps Program
• SBA SBIR/STTR Program Overview
• NSF SBIR Events and Bootcamps
• SBIR.gov Award Data