SBIR and STTR fund the same kind of R&D -- early-stage technology development that the government wants to see commercialized. The difference is structural: who does the work, where the PI sits, and whether you need an academic partner. Most content online explains the rules. This guide helps you decide.
The core difference in one paragraph: SBIR requires the small business to perform at least 67% of Phase I work with the PI employed by the company. STTR requires a formal research institution partner -- a university, national lab, or nonprofit -- performing at least 30% of the work. Choose STTR if your technology depends on academic facilities or expertise you can't replicate in-house. Choose SBIR if your team can do the work independently.
Choosing the wrong track doesn't just mean a rejected application. It means 40-80 hours of proposal work pointed in the wrong direction. The five questions below will get you to the right answer in under 10 minutes.
What Is the Difference Between SBIR and STTR?
The SBIR vs STTR difference comes down to team structure and work distribution, not what gets funded. Here are the six dimensions that actually affect your decision:
| Dimension | SBIR | STTR |
|---|---|---|
| Work distribution | Company performs >= 67% of Phase I work | Company >= 40%, research institution >= 30% |
| PI employment | PI must be employed >= 51% by the company | PI can be employed by the research institution |
| Academic partner | Not required (can subcontract to universities) | Required -- must have a formal cooperative agreement |
| Subcontracting flexibility | Up to 33% to subcontractors/consultants | Remaining 30% can go to either party or third parties |
| Award amounts | Same as STTR within each agency | Same as SBIR within each agency |
| Agencies offering both | All 11 SBIR agencies | All 11 agencies also offer STTR |
Two things to notice here.
First, the PI employment rule is the biggest differentiator. If your lead researcher is a university professor who won't leave their faculty position, SBIR is off the table -- your PI needs to be on your payroll at 51%+ time. STTR was designed for exactly this situation: the PI stays at the institution, the company manages the commercial pathway.
Second, work distribution matters more than it looks. SBIR's 67% threshold means your company needs the in-house capacity to do two-thirds of the technical work. If your R&D depends on university labs, specialized equipment, or institutional expertise, hitting 67% in-house may require capabilities you don't have. STTR's 40% threshold is significantly more achievable for early-stage startups working alongside academic partners.
How to Choose Between SBIR and STTR: 5 Questions
Answer these five questions and you'll know which track fits.
Question 1: Is your PI employed full-time (51%+) by your company?
If yes: SBIR is available to you. Move to Question 3.
If no: Your PI doesn't meet the SBIR employment threshold. You need STTR -- or you need to restructure your team so the PI is on your company's payroll at majority time.
Why it matters: This is the most common disqualifier in SBIR applications. Founders assume their university collaborator can serve as PI on an SBIR. They can't. The PI must be "primarily employed" by the small business concern (SBC), which every agency defines as >= 51% time. If your lead researcher holds a faculty position and can't commit majority time to your company, SBIR is not available to you without a team restructure.
Question 2: Does your technology require academic labs, equipment, or specialized facilities?
If yes: STTR gives you a natural structure to access those resources through the required research institution partnership. With SBIR, you'd need to subcontract (capped at 33%) and still perform 67% in-house.
If no: Either track works mechanically. Move to Question 3.
Why it matters: A biotech startup developing a novel assay that requires a BSL-3 lab isn't going to build one. A materials science company that needs a synchrotron beam line can't buy one. STTR's structure assumes the company doesn't have every resource -- that's why the research institution exists in the arrangement.
Question 3: Is your PI a US citizen or permanent resident?
If yes: All agencies are open to you. Move to Question 4.
If no: You cannot apply to NSF -- period. NSF is the only major SBIR/STTR agency that requires the PI to be a US citizen or permanent resident. NIH, DOD, DOE, and NASA do not have this requirement. If your lead researcher is on an H-1B or other work visa, target NIH or DOD programs instead.
Why it matters: This catches international founders every cycle. NSF's SBIR program is otherwise one of the most founder-friendly -- $305K Phase I awards, clear evaluation criteria, rolling deadlines for some programs. But the citizenship requirement is absolute. No exceptions, no waivers. Founders who discover this after spending weeks on an NSF application have lost that time entirely.
Question 4: Can your company perform 67% of the Phase I work in-house?
If yes: SBIR is the simpler path. No partner negotiation, no cooperative agreement, no splitting IP discussions with a university tech transfer office.
If no: STTR's 40% company threshold is more realistic. If your in-house team can handle 40-50% of the technical work but needs an academic partner for the rest, STTR is the structurally honest choice.
Why it matters: Reviewers can tell when a company claims to do 67% of the work but the technical plan clearly requires institutional resources. Overstating your in-house capacity doesn't just risk a compliance issue -- it undermines your proposal's credibility with peer reviewers.
Question 5: Do you have or can you establish a university research partnership?
If yes and STTR is indicated: You're set. Budget time for the cooperative agreement negotiation -- typically 4-8 weeks with a university tech transfer office.
If no and STTR is indicated: You need to find a partner. This is often easier than founders expect. University faculty actively look for STTR partnerships because it funds their research. Start with professors who've published in your technology area, or contact the university's Office of Sponsored Research. Cada's eligibility assessment includes partner sourcing guidance.
If SBIR is indicated: You can still collaborate with universities as subcontractors (up to 33% of the budget) without the formal STTR cooperative agreement.
Decision Summary
| Your Situation | Recommended Track |
|---|---|
| PI is your employee, work is in-house | SBIR (simpler, no partner overhead) |
| PI is at a university | STTR (required -- PI can't serve on SBIR) |
| Need academic labs/equipment for R&D | STTR (structured for institutional access) |
| Non-citizen PI targeting NSF | Neither at NSF -- apply NIH or DOD instead |
| Can't do 67% in-house, have academic partner | STTR (40% threshold is more realistic) |
| Could go either way | SBIR preferred (less overhead), unless STTR adds strategic value |
SBIR vs. STTR Rules by Agency
The decision framework above covers federal rules that apply everywhere. But agencies add their own wrinkles. Here are the ones that change your calculus:
| Agency | Phase I Amount | PI Citizenship | Funding Type | Key Difference |
|---|---|---|---|---|
| NIH | Up to $314,363 | Not required | Grants (R43/R41) | Largest SBIR funder. Study section review. Same success rate (~20-25%) for both tracks. |
| NSF | $305,000 | Required (US citizen/PR) | Grants | Most restrictive PI rule. Otherwise very founder-friendly. |
| DOD/AFWERX | $75K (SBIR) / $110K (STTR) | Not required | Contracts (not grants) | 90-day Phase I timeline. STTR awards $35K more. FOCI/ITAR screening. |
| DOE | Varies by office | Not required | Grants | Office of Science vs ARPA-E have different processes. Recently rescinded indirect cost caps. |
| NASA | Varies | Not required | Grants | Topic-driven solicitations with annual BAA cycles. |
| USDA | $100-150K | Not required | Grants | NIFA-administered. Smaller program (~$25M annual budget). |
Three Agency-Specific Gotchas
NIH's funding mechanism codes tell you the track. R43 = SBIR Phase I. R41 = STTR Phase I. R44 = SBIR Phase II. R42 = STTR Phase II. When searching NIH Reporter for comparable awards, filter by these codes to see which track similar companies used.
DOD STTR pays more than DOD SBIR at Phase I. AFWERX awards $110K for STTR Phase I vs. $75K for SBIR. If you have a research institution partner and you're targeting defense, STTR is worth $35K more at Phase I -- a meaningful difference for a 90-day project.
NSF's citizenship requirement applies to both SBIR and STTR. Some founders assume the STTR track bypasses the citizenship rule because the PI can sit at a university. It doesn't. NSF requires US citizenship or permanent residency for the PI regardless of track.
3 Mistakes That Cost Founders Months
Mistake 1: Choosing SBIR by default when STTR would be stronger
A computational biology startup has a CTO running the company and a collaborating professor running the experiments. The CTO files an SBIR application listing themselves as PI and subcontracts 30% of the work to the professor's lab.
The problem: the CTO's technical background is in software, not wet lab biology. Reviewers score the "Investigator" criterion poorly because the PI lacks domain expertise in the proposed experiments. An STTR with the professor as PI would have scored significantly better on investigator qualifications -- and the 40/30 work split would have honestly reflected how the research actually gets done.
Mistake 2: Discovering the PI employment rule after writing the proposal
An AI diagnostics startup recruits a university researcher as their PI for an NIH SBIR. They write the full Specific Aims page and begin the Research Strategy. During budget preparation, they realize the PI needs to be employed >= 51% by the company during the award period. The PI is a tenured professor and can't leave for majority time.
The fix requires switching to STTR, rewriting the budget, negotiating a cooperative agreement with the university's tech transfer office (6-8 weeks), and restructuring the work plan. If the receipt date is in 3 weeks, this proposal isn't making it.
Mistake 3: Assuming NSF allows non-citizen PIs because NIH does
A medical device startup with a CTO on an H-1B visa prepares an NSF SBIR application. They've heard that "SBIR doesn't require citizenship" -- which is true for NIH, DOD, DOE, and NASA. NSF is the exception. The application is administratively rejected before peer review.
Had they targeted NIH NIBIB (National Institute of Biomedical Imaging and Bioengineering) instead, the same technology and the same PI would have been eligible. The proposal work wasn't wasted -- it needed to be reformatted from NSF's structure to NIH's -- but the time pressure of missing the NSF cycle and catching the next NIH receipt date (April 5, August 5, or December 5) cost them months.
Frequently Asked Questions
Can I switch from SBIR to STTR after submitting?
No. SBIR and STTR are separate tracks with separate funding announcements. You submit to one or the other. If you realize mid-writing that you're on the wrong track, switch before submission rather than submitting to the wrong one.
Does STTR cost more because of the university partner?
Sometimes. Universities charge indirect costs (F&A rates) that can range from 15% to 60%+ of direct costs. This doesn't change your total award amount -- the agency funds the full budget -- but it means more of your budget goes to institutional overhead and less to direct research activities. Negotiate the indirect rate early. Some universities will accept a reduced rate for STTR proposals.
Can I apply for both SBIR and STTR at the same agency simultaneously?
Yes, but only for different projects. You cannot submit the same research idea as both an SBIR and an STTR to the same agency. You can submit an SBIR for Project A and an STTR for Project B to the same agency in the same cycle.
What if I want STTR but don't have an academic partner yet?
Start with your technology domain. Search for faculty at research universities who publish in your area -- they're often interested in STTR partnerships because it funds their research with minimal PI time commitment. Contact the university's Office of Sponsored Research or Technology Transfer office. Cada's eligibility assessment includes partner sourcing guidance for founders who need to establish these relationships.
Can I apply for SBIR at one agency and STTR at another for the same technology?
Yes. This is a smart portfolio strategy. You might submit an SBIR to NIH (where your internal team leads) and an STTR to NSF (where a university partner adds credibility in a different technical dimension). Cross-agency applications are independent and reviewed separately.
Not Sure Which Track Fits Your Startup?
The framework above covers the structural decision. But the right track also depends on your specific technology, your team's background, the agency's current priorities, and your timeline.
Cada has written 100+ proposals across 30+ agencies -- including both SBIR and STTR applications at NIH, NSF, DOD, DOE, and NASA. The most common mistake we see is founders committing 40+ hours to a proposal on the wrong track.
If you're not sure whether SBIR or STTR is the right fit, Cada's free 15-minute eligibility assessment covers track selection, agency matching, and academic partner sourcing guidance. No pitch, no obligation -- just a straight answer on where your company is competitive.