If you're pursuing federal funding, you're choosing between SBIR and STTR. Both fund technology development without taking equity. Both run in three phases. Both connect startups to federal agency missions. But they fund different company structures, require different partnerships, and suit different paths to commercialization. Pick the wrong one and you lose time, proposal effort, and a shot you won't get back.
I have spent years walking founders through non-dilutive funding inside federal agency innovation programs and I-Corps customer discovery. Below is how I explain the difference between SBIR and STTR, how I help founders pick, and what I keep seeing get applications rejected.
The fundamental difference: research partnerships and eligibility
SBIR lets you do all the research and development work in-house. You control the technology, hire the team, and manage the project without any external research partnership.
STTR requires a formal research partnership between your company and a qualified research institution, usually a university or nonprofit research lab. At least 30 percent of the work has to come from the research partner, and at least 40 percent from you. The remaining 30 percent can go to either party or a subcontractor.
That one structural difference decides most of the choice for the founders I work with. If you already have the technical capability in-house and don't need university research infrastructure, SBIR gives you more control and fewer partnership complications. If your technology depends on university labs, specialized equipment, or faculty expertise, STTR gives you the funding and the structure to formalize that collaboration.
Who qualifies for each program
Both programs require you to be a for-profit small business, U.S.-owned and operated, with fewer than 500 employees including affiliates. Your principal investigator has to be primarily employed by your company at the time of award and for the life of the project.
SBIR has no partnership requirements beyond standard subcontracting rules. You can work alone or bring in subcontractors as needed.
STTR requires a formal cooperative research and development agreement with a nonprofit research institution before you apply. That partner has to be a university, a federally funded research and development center, or a nonprofit research organization, and the partnership needs to show real collaboration, not just a university's name attached for credibility.
Phase structure and typical award amounts
Both programs run in three phases. Phase I funds feasibility studies and proof of concept. Phase II funds full research and development toward a working prototype or product. Phase III is commercialization, pursued through non-SBIR sources like contracts, venture capital, or revenue.
Phase I awards typically run from $50,000 to $275,000 over six to twelve months, depending on the agency. Phase II awards run from $750,000 to $2 million over two years. Some agencies offer a Direct to Phase II path for companies with strong preliminary data, so you can skip Phase I and compete straight for Phase II funding.
Award amounts vary by agency mission and topic. Defense and intelligence agencies tend to fund at the higher end. Health and energy programs vary more. The Small Business Administration SBIR/STTR resources page has agency-specific details and historical award data.
What agencies actually buy, and why proposals get rejected
The applications I see fail usually pitch technology without showing mission fit. Agencies fund solutions to specific mission problems they cannot solve with what they already have, and interesting science on its own does not clear that bar.
Before I let a founder write a proposal, I make them answer four questions. What mission problem does the agency face? Which program office owns that problem and controls the budget? What's the current solution, and why does it fail? How does your technology close a capability gap the agency can't close internally?
I treat customer discovery with agency stakeholders as mandatory, and my read on what agencies buy comes partly from sitting on the NIST Manufacturing Extension Partnership advisory board. Interview program managers, end users, and technical leads to validate the problem, understand procurement constraints, and learn what evidence would convince the agency to move you from Phase II into operational procurement. I hold agency interviews to the same bar I set in how many customer interviews you actually need: a repeated pattern across many conversations, not one friendly chat with a single program manager. The discipline is the same one behind customer discovery without a paying customer. The agency is not writing you a check during Phase I, but validating that they actually own the problem is still real evidence work, and skipping it is how a technically strong proposal loses anyway.
Proposals that show mission alignment, cite real conversations with agency stakeholders, and lay out a commercialization path beyond the award itself win at a higher rate than proposals pitching generic technology capabilities. Presenting to a program office has more in common with pitching any institutional buyer than with pitching a venture investor: you are selling to someone who has to justify the decision upward, not just defend the mission case on paper.
When I read rejection feedback, the same reasons come up again and again. The technology doesn't align to the agency topic or mission priorities. The team lacks relevant experience or a track record delivering similar work. The proposal solves a problem the agency doesn't prioritize. There's no evidence of customer discovery with agency stakeholders. The commercialization plan is vague or ignores anything beyond federal contracts.
Technical reviewers also reject proposals with weak Phase II transition plans, budgets that underestimate development complexity, and teams missing skills the work actually requires.
I tell founders they improve their odds by aligning the proposal to published agency topics, showing prior customer discovery with the target agency, building a team with relevant domain expertise and past performance, laying out a realistic commercialization path that includes dual-use commercial applications, and budgeting honestly instead of stretching thin to hit a funding cap.
STTR-specific challenges: making partnerships work
STTR brings partnership complexity that SBIR skips entirely. You and the research institution have to negotiate intellectual property rights before you submit the proposal. Work allocation between the two of you has to be realistic and defensible. Communication and project management across institutional lines need real structure.
The STTR partnerships I have seen work spell out who owns background IP, who controls foreground IP developed during the project, and how commercialization rights get managed. They set up regular coordination, clear decision authority, and a way to resolve conflict before problems show up.
I always check whether the university research partner has real experience with SBIR/STTR projects and can move at the pace federal contracts demand. Academic timelines don't always match startup urgency or federal reporting requirements.
Which program fits your startup
I steer founders to SBIR when they have technical capability in-house, want full control over IP and execution, don't need university research infrastructure, and can staff the project with internal hires or independent contractors.
I steer them to STTR when the technology depends on university labs or specialized equipment, if faculty or research staff bring expertise you can't hire, if your commercialization path benefits from university partnership credibility, or if you need the formal structure to turn an informal university collaboration into funded work.
Some startups pursue both, applying for SBIR when the work can be done internally and STTR when a specific project needs a university research partner. The programs are not mutually exclusive, though each application still has to meet its own structural requirements.
Frequently asked questions
Can I apply for both SBIR and STTR at the same time?
Yes. The two programs are not mutually exclusive, and I have coached founders who ran an SBIR proposal for the piece of the technology they could build in-house alongside an STTR proposal for the piece that needed a university lab. Each application still has to clear its own eligibility and partnership requirements on its own terms.
Do I need an advanced degree or a research background to apply for STTR?
No. You need a formal partnership with a qualified research institution, not a personal research background. Your principal investigator has to be primarily employed by your company, and the research partner supplies the technical depth you don't have in-house. I have worked with founders whose only academic credential was an undergraduate degree, paired with a faculty co-investigator who filled the research gap.
What happens if my STTR research partner drops out during the project?
You have to notify the agency and typically find a replacement partner that meets the same eligibility rules, since STTR funding depends on that formal partnership existing for the life of the project. I tell founders to negotiate an exit clause into the cooperative agreement before signing it, because I have watched a project stall for months after a faculty partner changed institutions and nobody had planned for it.
Does taking venture capital disqualify a company from SBIR or STTR?
Taking venture capital does not disqualify a company outright, though ownership rules apply. A company owned by multiple venture capital firms can still qualify at some agencies under SBA's alternate size standard, and the specific ownership and control thresholds vary by agency and change periodically. Confirm the current SBA size and ownership rules before you assume either program is open or closed to you.
Is SBIR or STTR harder to win?
STTR usually draws fewer applicants per topic than SBIR, since the partnership requirement filters out founders who don't want the added coordination, so the odds per application can run better. SBIR draws more applicants per topic, but you also have more agencies and more topics open to you, since you're not limited to opportunities that match a specific research partner's capability.
My technology needs some outside expertise but not a full lab partnership. Which program fits?
Start with an honest inventory of what you can build without outside help. If the gap is a capability you cannot hire or buy your way into within the award timeline, that gap points you toward STTR and the specific institution that has it. If the gap closes with the award money itself, contracted help, or a hire, SBIR keeps you in full control of the outcome.
Wherever you land, start with customer discovery focused on one specific agency mission problem. Identify the program office, the budget line, and the decision-maker who owns that problem. Validate that switching to your solution is worth the organizational friction. Then pick the SBIR or STTR path that matches your team structure and development needs.