NSF Seed Fund (SBIR/STTR) Phase I

NSF Seed Fund Phase I is one of the cleaner non-dilutive paths for a US startup or small business that is trying to turn a real technical breakthrough into something fundable, testable, and commercially credible. It is not a general startup grant and it is not a place to ask for broad product development money. The program exists for high-risk R&D where the technical uncertainty is real and where success could unlock a product, platform, or process with meaningful commercial and societal value.

The easiest way to think about it is this: NSF wants to fund work that is hard enough that ordinary operating revenue would be a weak fit, but focused enough that Phase I can make a clear go/no-go decision. If your project needs proof that a new material works, that a novel algorithm is materially better than current methods, or that a new hardware approach can survive a difficult technical test, this program may fit. If your main need is hiring a product team, polishing a demo, or scaling sales, it probably does not.

The official process starts with a Project Pitch. If NSF likes the fit, you get an invitation to submit a full proposal. Only invited full proposals are reviewed. That invitation gate is the key feature teams miss when they are new to NSF: you do not just prepare a proposal and send it in whenever you are ready. You first show that the project belongs in NSF Seed Fund, then you use the invite to enter the full-proposal process through Research.gov.

At a glance

What this program is really for

The Seed Fund exists to back technology that is still uncertain enough to deserve R&D funding. That can mean the science is promising but not yet proven, the engineering path is novel, or the commercialization path depends on solving one or more hard technical problems before the market can be addressed. NSF is not looking for polished marketing language. It is looking for a convincing technical story.

That story should connect three things:

1. A clear technical problem or opportunity.
2. A novel approach with genuine technical risk.
3. A plausible commercial outcome if the risk is retired.

This is why the best applications usually sound more like a research-and-translation plan than a standard startup pitch deck. The company still needs commercial intent, but the proposal has to show that commercialization depends on solving a technical problem, not just on branding or distribution.

The current solicitation also makes clear that the program is meant for broad science and engineering innovation. NSF is comfortable with deep-tech work across software, hardware, materials, devices, robotics, sensing, advanced manufacturing, data systems, and other technically demanding areas. The important question is not the sector label. It is whether the work has enough technical novelty and uncertainty to qualify as R&D.

What it offers

For Phase I, the current solicitation allows requests of up to $305,000 for a period of 6-18 months. That is meaningful support, but it is still a bounded proof phase. You should use it to reduce uncertainty, not to assume the company can finish every follow-on task inside one award.

Two practical details matter here:

• NSF expects teams to think about commercialization support inside the budget. The solicitation says Phase I recipients should budget for Technical and Business Assistance (TABA) and I-Corps training.
• The solicitation also says that other supplements to Phase I awards will not be allowed.

So if you are deciding whether the opportunity is worth the time, do not only look at the headline amount. Ask whether your team can use 6-18 months to answer the one or two technical questions that matter most for the next financing, customer, or partnership decision. If the answer is yes, this is the kind of program that can de-risk a deep-tech company without dilution. If the answer is no, the award may be too specialized to justify the compliance load.

Who should apply

This program is usually worth serious attention if you are a small business with a technical team, a real product thesis, and a hard problem that has not already been solved in the market. Strong candidates tend to be teams that can explain why the current approach is inadequate and why their own solution is materially different from existing options.

You are a better fit if:

• the project depends on genuine technical uncertainty;
• your company can explain why the work is hard, not just why it is desirable;
• you have a credible plan for how Phase I will reduce risk;
• you can show why the result matters commercially if the technical work succeeds;
• you are prepared to follow a federal proposal process with registrations, forms, and deadlines.

You are probably not a good fit if:

• the main need is customer acquisition, branding, or generic product work;
• the proposal is really asking for implementation of a well-understood solution;
• the team cannot identify a clear technical risk;
• the company does not meet the ownership, size, or PI requirements;
• you are not ready to handle federal compliance steps in a short window after an invitation.

Another useful way to think about it is this: NSF Seed Fund is not looking for the biggest company. It is looking for the most credible technical problem. A tiny team with unusually strong evidence and a sharp R&D hypothesis can be competitive if the proposal is disciplined.

Eligibility

The official get-started page gives the core requirements that matter most before you spend time on a pitch:

• the company must be a small business with fewer than 500 employees;
• the company must be located in the United States;
• at least 50% of the company’s equity must be owned by US citizens or permanent residents;
• NSF does not fund companies that are majority-owned by multiple venture capital firms, private equity firms, or hedge funds in SBIR/STTR;
• all funded work, including work done by consultants and contractors, must take place in the United States;
• the principal investigator (PI), described by NSF as the tech lead, must be legally employed at least 20 hours a week by the company;
• the PI does not need an advanced degree, but does need to commit the required project effort;
• the full proposal generally requires an official Project Pitch invitation.

The solicitation adds a few more caveats worth knowing early. NSF SBIR/STTR does not support clinical trials, and it does not support proposals whose commercialization path involves the production, distribution, or sale of certain Schedule I controlled substance-related chemical components or derivatives. If your company is in a regulated or sensitive area, read the solicitation carefully before investing in the pitch.

If you are choosing between SBIR and STTR, the Seed Fund pages also explain that the routes are similar but not identical. STTR requires a formal research partner arrangement and different work-sharing rules. SBIR is the more straightforward path if your company is doing the core work itself. If you are unsure which route applies, that is a good signal to review the solicitation before writing.

How to apply

The official sequence is simpler than many federal programs, but it still has a few moving parts.

1. Check fit before you write

Start by testing whether the project is genuinely an NSF Seed Fund project. Ask:

• What is the technical hypothesis?
• What is the hardest part?
• What would prove the idea is viable?
• What commercial decision depends on that proof?

If those questions are hard to answer in plain English, the application is not ready yet.

2. Submit a Project Pitch

The Project Pitch is the first formal gate. NSF says you can submit it anytime, and the page indicates you should hear back in about 1-2 months. The pitch should explain the project objectives, the technical innovation, and the technical risks. It is not the place to pad the narrative with buzzwords. It is the place to show that the work is both novel and hard.

Keep the pitch focused. A reviewer or program officer should be able to understand within a few minutes:

• what the company is building;
• what makes it technically risky;
• why the approach is different;
• what the commercial opportunity is;
• why NSF funding, not standard operating funds, is the right tool.

3. If invited, prepare the full proposal

If NSF invites you, you then prepare the full Phase I proposal. The full proposal must match the project described in the pitch. This is not the time to quietly pivot to a different product idea or expand the scope dramatically.

The full proposal is submitted through Research.gov, not by email and not through the pitch system. NSF says proposals are due by 5:00 PM submitter’s local time on the stated deadline. That detail matters because the deadline is firm; do not assume a grace period.

4. Submit within the valid window

The solicitation says invited proposers must submit within the next two deadlines after the invitation date. That means the invitation is useful, but it is not open-ended. You should be ready to submit soon after the pitch is approved, not months later after the team has moved on to other work.

Timeline and deadline reality

The current Seed Fund pages show a process that is partly rolling and partly deadline-driven.

The pitch stage is continuous: NSF says you can submit a Project Pitch anytime. The response time is roughly one to two months, though that is not a promise of approval.

The full proposal stage is deadline-based. The full-proposal page says the submission deadlines are coming soon, while the NSF solicitation page I checked lists deadlines through November 5, 2025. Because the current date is later than that, you should not assume a new window without checking NSF directly. The safe rule is to verify the live NSF page before you plan staffing or budget timing around a specific due date.

For practical planning, the key point is that the invitation-to-deadline window can be short. That means you should not wait until you receive the invite to start organizing the materials you will almost certainly need. Build the compliance pieces early, even if the pitch is still under review.

Required materials and prep work

The exact full-proposal package depends on the solicitation and project structure, but the official pages make several items clear enough to prepare in advance.

Expect to need:

• a tight project description tied to the pitch;
• a budget that fits the Phase I cap and duration;
• registration in the required federal systems;
• proposal documents prepared in the NSF format;
• biographies and support statements for key personnel;
• collaborator and affiliation information;
• commercialization-related justification if you budget for TABA or I-Corps.

The solicitation specifically says that Biographical Sketch(es), Current and Pending (Other) Support, Collaborators and Other Affiliations (COA), and Synergistic Activities must be submitted for the PI and other key personnel as applicable. It also says the biosketch and current/pending support forms must be done in SciENcv.

There are also registration steps that can take time:

• SAM registration;
• Research.gov company registration;
• SBIR Company Registry registration.

If you are a first-time applicant, these are not trivial admin steps. Build time for them early. A common mistake is waiting until the invite arrives and then discovering that one registration or profile is missing, which compresses the submission window and forces rushed work.

How to decide whether it is worth your time

The program is worth the effort when it does at least one of these things for your company:

• proves the core science or engineering works;
• reduces a major technical risk that investors or customers care about;
• creates evidence that your approach is meaningfully better than alternatives;
• unlocks a follow-on Phase II path, strategic partnership, or fundraise;
• helps your company reach a credible inflection point without dilution.

It is probably not worth the effort if you cannot articulate a hard technical milestone. NSF reviewers do not reward vague optimism. They want a project that can be evaluated on substance. If your only goal is to say you won an NSF grant, this is the wrong reason to apply. If your goal is to answer an expensive technical question that your company must solve anyway, it may be one of the best options available.

In other words, the right question is not “Can we use the money?” It is “Can we use the money to retire the most important technical uncertainty in the business?”

What a strong proposal narrative looks like

A strong NSF Seed Fund proposal is usually easy to summarize in one or two sentences. It says what problem exists, why current approaches are limited, what your team will test during Phase I, and how the result will influence the company’s next move. The language can be technical, but it should still be readable by someone who knows the field and wants a clear argument.

One useful structure is:

• the market or scientific need;
• the technical barrier that keeps existing solutions from solving it;
• your proposed approach;
• the key experiment, prototype, or validation step in Phase I;
• the decision that follows from success or failure.

This structure helps because NSF review is not about rewarding the longest application. It is about rewarding a clean connection between uncertainty, method, and outcome. If the project can be explained as “we will test whether this genuinely hard thing works, and if it does, the company can move to the next stage,” the proposal is much easier to defend.

It also helps to keep the commercialization language concrete. Instead of saying the work will “revolutionize the industry,” explain what kind of customer, partner, or deployment becomes possible if Phase I succeeds. Concrete downstream value is more persuasive than big claims.

Tips that usually improve the odds

Keep the technical risk narrow

The best Phase I proposals usually identify one central technical risk and show how the project will measure progress against it. A proposal that tries to solve five problems at once often reads like a wish list. A proposal that defines one difficult problem and a clean test plan reads like real R&D.

Show why your team can execute

NSF wants novelty, but it also wants a team that can make progress. Explain why your founders, researchers, engineers, advisors, or partners are the right people for this particular risk. Prior experiments, prototypes, published work, prior deployments, or domain-specific experience all help.

Tie the commercial story to the technical one

Commercialization should not be separate from the research story. The best applications explain how the technical result changes the business decision. For example: if the prototype passes, the company can move to customer pilots; if the test fails, the company avoids a costly dead end.

Budget for the real process

This program includes federal compliance and proposal overhead. Do not build a proposal around an imaginary “lightweight” submission process. Budget time for registrations, internal review, budget validation, and document collection.

Treat the pitch as a decision tool

The pitch is not just a formality. It is the tool NSF uses to decide whether the full proposal is worth reviewing. If the pitch is weak, the rest of the process never starts. That makes the pitch one of the highest-value documents to get right.

Common mistakes

The most common error is writing a startup pitch instead of a research pitch. NSF already knows that founders believe in their product. What they need to see is a technical case that can survive scrutiny.

Other frequent mistakes include:

• treating the Project Pitch as optional;
• ignoring ownership or employment requirements;
• assuming a generic “AI,” “climate,” or “biotech” label is enough;
• failing to explain the technical risk in measurable terms;
• over-scoping the work beyond what Phase I can realistically prove;
• waiting too long after the invitation to start the full-proposal package;
• using stale deadline assumptions instead of checking the live NSF page.

The strongest way to avoid these mistakes is to write the proposal as if an informed technical reviewer will ask, “What exactly is uncertain here, and how will this project reduce that uncertainty?”

FAQ

Is this a loan or equity deal?

No. NSF Seed Fund Phase I is non-dilutive grant funding.

Do I need to submit a full proposal first?

No. You start with a Project Pitch, and only invited projects move to the full proposal stage.

Can I apply whenever I want?

You can submit a Project Pitch anytime, but the full proposal must fit an active NSF submission window and the invitation rules in the solicitation.

Is this only for very early-stage companies?

Not necessarily. The real test is whether the project has unresolved technical risk and a credible commercialization path. Some later-stage companies still fit if the project is genuinely R&D-heavy.

Can I use the award for ordinary product work?

Not as the main purpose. The program is for high-risk technology development and commercialization de-risking, not routine product backlog work.

What if I am unsure whether I should choose SBIR or STTR?

Read the solicitation and the eligibility guidance carefully. If you are still uncertain, the official pages point applicants to program staff and the proposal guidance resources.

Official links

• Program landing page: https://seedfund.nsf.gov/apply/get-started/
• Full-proposal guidance: https://seedfund.nsf.gov/apply/full-proposal/
• Submission instructions: https://seedfund.nsf.gov/how-to-submit/
• NSF solicitation: https://www.nsf.gov/funding/opportunities/sbirsttr-phase-i-nsf-small-business-innovation-research-small-business/nsf24-579/solicitation