SBIR and STTR university partners: the rules, how to find one, and what to agree before you submit

An STTR proposal requires a research-institution partner, usually a university, that does at least 30% of the work while the small business does at least 40%. An SBIR proposal doesn't require one, but it lets subcontractors and consultants, a university among them, do up to a third of the work in Phase I and up to half in Phase II, and 46% of SBIR Phase II recipient respondents in a National Academies survey reported a university connection of some kind. Whichever program you apply to, the partner is best found through its research record rather than its reputation, and four things are worth settling before you submit: the share of the work and the budget, who the principal investigator is, how intellectual property is handled, and the university's internal deadline.

This guide is for both sides of that conversation. It's written first for founders and PIs at small companies looking for a partner, and also for the faculty member who's been asked to join and the research office that has to sign. In the Helikon Method, finding a partner is our Match phase and agreeing the terms is Unite, and with SBIR and STTR both have to happen before the proposal goes in, not after the award.

One note on status before the rules. SBIR and STTR authority expired on 30 September 2025, and the programs were reauthorized on 13 April 2026, through 30 September 2031, by the Small Business Innovation and Economic Security Act. The 2026 law didn't change the rules this guide covers (the work split, the PI rules, the research-institution definition, and the STTR rights agreement), but agencies are rewriting their solicitations to match it, so check the current one before you plan around any figure here.

What's the difference between SBIR and STTR?

Both programs fund small businesses, not universities, to do research and development with a path to commercial use, and in both the small business is always the applicant. The difference is how much of the work a research institution does, and whether it has to be involved at all.

	SBIR	STTR
Research institution partner	Optional	Required
Minimum share performed by the small business	Two-thirds in Phase I, one-half in Phase II	40%
Minimum share performed by the research institution	None	30%
Can the agency allow a different split?	Yes, with written approval	No, the split is set by statute
Where the principal investigator works	Primarily at the small business	At the small business or at the research institution (agency rules vary)
Written IP agreement with the partner	Terms go in the subaward	An allocation of rights agreement is required before award
Participating agencies	11	6
Share of agency extramural R&D set aside	3.2%	0.45%

The rules come from the SBA's SBIR and STTR Policy Directive, and the agency counts and set-asides from the Congressional Research Service. STTR runs at USDA, Defense, Energy, HHS (including NIH), NASA, and NSF. SBIR runs at those six and at Commerce, Education, Homeland Security, Transportation, and EPA.

The set-asides explain why STTR is the smaller program. In FY2022, agencies obligated $4.4 billion through SBIR and $662.3 million through STTR. NSF's current solicitation, for example, estimates about 180 SBIR Phase I awards per year against about 50 STTR Phase I awards (NSF 26-510). Award sizes are set by each agency within SBA's guidelines. As of April 2026, an agency can make a Phase I award of up to $323,090 and a Phase II award of up to $2,153,927 without a waiver (SBIR.gov), and some set their own lower caps. NSF's Phase I, for instance, is up to $305,000, inclusive of indirect costs, fee, and TABA.

Phase I awards are smaller in amount, but the evidence shows that the award outcomes deliver more to small businesses than just the funding alone. In a study of Department of Energy SBIR applicants that compared companies just above and just below the award cutoff, Howell (2017) found that an early-stage award roughly doubled the probability of later venture capital funding, with large positive effects on patents and revenue, and the evidence points to the grant paying for prototype or proof-of-concept work as the most likely reason. That's a good reason to choose the program and the partner carefully rather than quickly.

Does an SBIR need a university partner?

No, but nearly half of SBIR projects report some university connection, and studies of outcomes suggest it helps.

The National Academies' assessment of the STTR program surveyed Phase II recipients at five agencies, and 266 STTR and 1,795 SBIR respondents answered its question on university links. Almost every STTR respondent (95%) reported a university connection of some kind, which you'd expect. But so did 46% of SBIR respondents. A research institution was a subcontractor on about a quarter of the SBIR projects (25.8%), faculty held a role other than PI on 26.0%, and graduate students worked on 20.3%. On the outcome side, a study of NIH SBIR awards by Link and Ruhm (2009) found that university involvement in the underlying research increased the probability of commercialization, and an analysis of SBIR firms by Siegel and Wessner (2012) found that start-ups with closer ties to universities performed better, with higher sales and more patents and licenses. The authors were careful to add that this could partly reflect universities choosing to work with the most promising firms.

So the useful question isn't "SBIR or STTR?" in the abstract. It's how much of the work a university needs to do, where the PI should sit, and which agency's topic fits. We cover that choice in detail in SBIR vs STTR: which program fits when you already have a university collaborator.

Translation Box: subaward

On the academic side: a sponsored-research agreement under a federal prime award. It runs through the sponsored programs office, carries the university's federally negotiated indirect cost rate, and has to comply with the federal terms that flow down from your award.

On the industry side: the contract you sign with the university so it can do its share of the work and get paid from your award. It feels like a vendor agreement.

In this guide: the academic meaning, which is why the university can't sign your standard vendor terms.

What counts as a research institution?

For STTR, the partner has to be a research institution as the policy directive defines it. That means an organization with a place of business in the United States that either operates primarily in the United States or makes a significant contribution to the U.S. economy, and that's either a nonprofit institution as defined in the Stevenson-Wydler Act or a federally funded research and development center (FFRDC). In practice, that covers universities, nonprofit hospitals, other nonprofit research organizations, and federally funded research and development centers (NIH Grants Policy Statement 18.5).

A for-profit lab, a university with no place of business in the United States, or an individual consultant can't be the STTR partner, although some of them can still join the project in other roles. NIH describes the STTR partner as "a single, non-profit research institution," so plan around one formal partner even when you'd like to involve two groups.

For SBIR, any qualified subcontractor or consultant can do the outside share, and they all draw on that same share, so a university is one option among several.

Who can be the principal investigator?

On an SBIR, the PI's primary employment has to be with the small business at the time of award and for the whole project. "Primary" means more than half of their employment time. A faculty member who's a co-founder can't be the SBIR PI while they're employed full time by the university, unless they reduce their appointment, and agencies grant deviations only rarely.

On an STTR, the policy directive allows the PI's primary employment to be with either the small business or the research institution, which is one of STTR's practical advantages for university-based researchers. Agencies add their own conditions, though, and they don't all agree:

• NIH lets an STTR PI work primarily at the research institution, but requires at least 10% of their time on the project and a formal appointment with, or commitment to, the small business.
• NSF requires the PI to be primarily employed (at least 51%) by the small business on both programs, and requires an STTR co-PI from the partner research institution.

Read the PI section of your specific solicitation before you write anyone's name on the cover page.

How is the work split measured?

Agencies measure the split either by total award dollars or by labor hours, and each solicitation says which. For STTR, the minimums are fixed by law and can't be waived, so a budget that puts the university at 29% doesn't get a deviation. It's noncompliant, and the agency can decline it on that ground. Build the budget to the rule with some margin, and agree with the university which costs, including their indirect costs, sit on their side of the line.

For SBIR, the two-thirds and one-half minimums can be adjusted with the funding officer's written approval, but treat that as an exception you have to justify rather than a plan.

How do you find a university partner?

Start from the problem, not the university's name. The best partner for a Phase I is rarely the most famous lab at the nearest large university. It's the group that has already measured what your specific aims propose to measure, has the equipment and students to do it inside a Phase I timeline, and has run a federally funded project before. Here's how to find that group:

• Search on your aims, not on institutions. Paste the problem you're solving, in the words your aims page uses, and look at who has published and patented on it. Look at the technical detail in their papers, not the department page.
• Read the funding history. A prior federal award on a related topic tells you the group can run a funded project. A paper that acknowledges an earlier SBIR or STTR award tells you they've been a small-business partner before, and they'll know the paperwork.
• Check capacity. Ask who would do the work. A Phase I that runs 6 to 18 months at NSF needs a student or postdoc with time, and equipment that isn't booked for a year.
• Check who they already work with. A group with a long relationship with one of your competitors may be the wrong partner, however good the science.
• Talk to the lab first, then the research office, and early. Faculty agree to collaborate. The sponsored programs office agrees to the terms, and it's the office that needs the lead time.

That last point is where first-time applicants often lose time. The National Academies survey found that almost three-quarters of STTR respondents already had a relationship with their research institution before the award. Most surveyed STTR projects built on a relationship that already existed, so if you're starting from zero, start well before the deadline.

If you're choosing between two or three candidate groups and want to perform rigorous due diligence on the decision, our Partner Evaluation Scorecard screens, scores, and ranks candidate partners on eight weighted criteria and prints a one-page decision memo you can share with a co-founder or an advisor.

Our technology area pages are one place to see how the field is distributed before you start. They show how many researchers, publications, and U.S. patents our Helikon Research Graph holds, and which universities, laboratories, and companies are most active in each.

What do you need to agree before you submit?

Everything a university needs to sign your proposal, plus the things that become disputes if you leave them until after the award. Our free SBIR and STTR partner readiness checklist puts all of it in one document. The core list:

• The statement of work and the split. What the university does, what you do, and how the split is measured under your agency's rule.
• The budget. The university's subaward budget, including its indirect costs. On a grant, the agency's terms usually apply the Uniform Guidance (NSF's solicitation, for example, requires pass-through entities to follow 2 CFR 200.332), which requires you to use a subrecipient's federally negotiated indirect cost rate where it has one, so the rate isn't yours to negotiate down. Plan the university's share with that rate in it. The university can choose to charge less, though, and some universities, or schools within them, have published reduced or waived indirect cost rates for Phase I SBIR and STTR proposals, to encourage more Phase I activity. Ask the research office whether it has one.
• The people. Who the PI is, who the co-PI or key person at the university is, how much of their time each commits, and whether the agency's PI rules are met.
• Intellectual property. For STTR, the allocation of rights agreement (next section). For SBIR, the IP terms in the subaward. Under the Bayh-Dole Act, both a small business and a university can generally elect to keep title to the federally funded inventions they own (35 U.S.C. 202), so expect the university to start from "each side owns what its people invent." The law only lets an organization keep what it already owns, though. The Supreme Court held in 2011 that Bayh-Dole doesn't automatically give an organization title to its people's federally funded inventions, and that the rights start with the inventor, who has to assign them (563 U.S. 776). So confirm that everyone working on the project on both sides, students and consultants included, is covered by their organization's invention assignment. Our free Contract Red Flags Kit lists the IP and publication clauses that slow these agreements down, with a balanced position for each.
• Publication. How long you get to review a paper before it's submitted, and whether you can delay it for a patent filing. The SBA model also gives the reviewing party a "right of refusal" it doesn't define, and universities will usually want that narrowed to removing your confidential information. They generally won't accept a broader veto.
• Conflicts of interest. If the faculty member holds equity in your company or is paid by it, their university's own policy will almost certainly require a disclosure. From NIH Phase II on, the federal conflict of interest rules also require the institution to review the interest and, if it's a conflict, manage it (NIH exempts Phase I SBIR and STTR). This is routine, but it takes time.
• The institutional commitment. Expect to need a letter of commitment or a signed subaward package from the university's authorized official, not just from the professor.

What goes in the STTR allocation of rights agreement?

Before an STTR award is made, the small business and the research institution must have a written agreement that allocates intellectual property rights and the rights to carry out follow-on research, development, or commercialization. The policy directive says the small business submits it with the proposal and certifies that it's satisfactory, so check your solicitation for exactly what it asks for and when. At a minimum, the agreement has to answer three questions:

1. Who's responsible for, and who owns, inventions that come out of the project, including expenses, liability, revenues, and profits?
2. Which party may file for patents, in the United States and abroad?
3. Which party may continue the research, including follow-on awards outside the STTR program?

The model agreement that SBA provides, and that agencies adopt, gives a sensible starting point. An invention belongs to the party whose people conceived it or first actually reduced it to practice. Joint inventions are jointly owned, with the small business holding the first option to protect them. The small business gets an exclusive option to license the university's inventions from the project, and if it extends that option, it reimburses the university's patent costs up to the end of the option period. The government keeps a nonexclusive license for its own purposes. Most of the defaults are balanced. The option periods, the royalties, and the revenue split are left to the parties, and those are the terms you'll negotiate.

Many universities have their own version of this agreement ready to go, and starting from theirs is usually faster than starting from yours. We explain each clause, and the questions founders usually get wrong, in The STTR allocation of rights agreement, explained for founders.

How long does the university need?

Longer than the professor will tell you, because the professor isn't the one who signs. Every university proposal policy we checked publishes an internal deadline ahead of the agency's, so its sponsored programs office has time to review the budget, the statement of work, and the institutional commitment. For a complete, final proposal, the published policies we found ask for three to five business days, and one asks for the administrative pieces five business days out and the final package two days out. Guidance written specifically for SBIR and STTR partners asks for more notice than that. One university's timeline wants a notice of intent 30 days before the company submits, and another asks its researchers to contact its sponsored programs office as soon as possible when a company plans to bring the university in as a subcontractor, because the process takes time. If an allocation of rights agreement or a conflict-of-interest review is involved, add time for those as well. Ask the office for its internal deadline in your first week, not your last.

For the professor's side of the same conversation, including what they'll ask for and what makes saying yes easier, see What to offer a professor when you ask them to join your SBIR or STTR proposal.

What are the most common mistakes?

• Asking the department chair instead of the lab. Chairs forward emails. Find the faculty member whose papers match your aims and write to them directly, with the aims attached.
• Leaving the research office until after the award. The office signs the commitment, and on STTR it negotiates the rights agreement. Bringing it in late is a common reason a willing professor turns into a missed deadline.
• Assuming the university will accept your IP terms. It won't sign a vendor agreement that assigns its inventions to you. Start from the model agreement or the university's template.
• Getting the STTR math wrong. At 40 and 30, there's only 30% left for everyone else, and STTR allows no deviations. Check the split under your agency's measurement before the budget is final.
• Naming the wrong PI. A full-time faculty member can't be an SBIR PI, and at some agencies can't be an STTR PI either.
• Missing the agency's first step. NSF, for example, requires a Project Pitch and an invitation before a Phase I proposal, and a response to a pitch typically takes one to two months. Its next deadlines are 4 November 2026, 4 March 2027, and 7 July 2027, so a pitch sent now realistically points at March.

What changed in 2026?

The reauthorization kept the structure and added three things applicants should notice:

• Proposal limits. The law requires every agency to cap how many proposals a small business can submit, counted per fiscal year, per solicitation, or per topic, by FY2027 at the latest. Some agencies started early. HHS, which includes NIH, now accepts at most nine Phase I, Fast-Track, and Direct to Phase II applications from one company per fiscal year.
• Wider foreign-risk due diligence. Agencies' due diligence now covers the foreign affiliations and business relationships of a company's covered individuals, owners, and other key personnel with individuals or entities in foreign countries of concern, and any company relationship with entities or individuals on eight named federal lists.
• A new SBIR "strategic breakthrough" Phase II. Agencies with more than $100 million per year in required SBIR spending may award up to $30 million to one company, in one award or a series over up to 48 months. The company needs at least one prior Phase II award and new matching funds of at least 100% of the award, and not every eligible agency has chosen to offer it.

None of these changes the university partner rules. The proposal limits do make the choice of topic, and partner, count for more.

What happens after the award?

The award starts a funded research collaboration between your company and the university, and it needs diligent project management. Agree on who decides what, how often the teams meet, how milestones are tracked, and how disagreements escalate. Our plain guide to managing a university-industry research collaboration covers all of it, and the free Governance Charter puts it in one document you can sign at kickoff.

A note on what we don't do. Helikon Labs doesn't write, submit, file, or review SBIR or STTR proposals. We help you find partners, agree on governance, and manage your projects. We're also not a law firm, and this guide isn't legal advice, so take the subaward and the rights agreement to your counsel and the university's research office.

Where should you start?

With your specific aims and a search. On the Helikon platform, you can paste your aims into Partner search, which matches researchers by their record of publications, patents, and profiles, and is unlimited on the free plan. Each researcher's Funding tab lists the awards they've published on, including small-business awards their papers acknowledged, so you can see who's been a partner before. When you've found the group you want to work with, message them or invite them from their profile, and set up the proposal as a project with the SBIR/STTR agreement type, its timeline, and its contract requirements and dates. The Funding tab in Discover also shows the SBIR and STTR solicitations that Grants.gov lists, and matches relevant funding calls to your profile.

If you'd like a second opinion from a senior advisor on your negotiating position before you send the subaward or the rights agreement, see Office Hours.