SBIR Success Rates by Agency: 2025-2026 Data and Trends

A 20% success rate sounds reasonable until you realize it means four out of five proposals get rejected. That is the approximate overall award rate for SBIR Phase I across all federal agencies — but the number obscures enormous variation. Submit the same quality proposal to NIH versus the Air Force, and your odds shift by 10 percentage points or more.

Understanding where those odds fall, and why, is the difference between a strategic submission plan and an expensive lottery ticket. With SBIR reauthorized through 2029 and agencies expected to release new solicitations starting in April 2026, now is the time to study the data.

The Agency-by-Agency Breakdown

The most competitive SBIR program in the federal government is NIH. Across its 27 institutes and centers, NIH Phase I success rates have hovered between 15% and 18% over the past three fiscal years. The National Cancer Institute and NIAID are particularly competitive, with some review cycles dipping below 14%. NIH receives more SBIR proposals than any other agency — over 3,500 Phase I applications in a typical fiscal year — and its dual-review system (scientific merit plus commercialization) creates a high bar.

NSF sits in the middle range at roughly 20-25% for Phase I. The Project Pitch pre-screening system filters out non-competitive submissions early, which inflates the success rate for full proposals but still means three out of four pitches are declined before the proposal stage.

DoD is where the numbers get interesting because each service branch runs its own SBIR program. The Army's Phase I success rate has averaged around 20%, the Navy approximately 18%, and the Air Force (now including Space Force topics) around 22-24%. DARPA's SBIR success rates are harder to pin down because many awards come through Broad Agency Announcements rather than standard SBIR solicitations, but funded proposals tend to cluster around 25-30% for well-matched topics.

The Department of Energy awards Phase I proposals at roughly 20%, though ARPA-E's SBIR subset trends higher given its smaller applicant pool and focused topic areas. USDA, EPA, and the Department of Education have smaller SBIR programs with success rates that fluctuate more year to year, generally ranging from 18% to 28%.

Data sources for these figures include SBIR.gov award databases, SBA annual reports to Congress, and agency-specific dashboards. Exact rates shift by fiscal year, topic, and component — the numbers above reflect three-year averages through FY2025.

What Actually Drives the Differences

Success rate variation across agencies is not random. Three structural factors explain most of it.

First, proposal volume relative to budget. NIH's SBIR set-aside generates roughly $1.2 billion annually, but the sheer number of proposals from biotech and medtech companies keeps competition fierce. DoD's combined SBIR budget exceeds $2 billion, but it spreads across hundreds of narrow topics — many of which attract fewer than 20 proposals each. A DoD topic with 15 submissions and 4 awards has a 27% hit rate. An NIH omnibus solicitation topic with 200 submissions and 30 awards sits at 15%.

Second, topic specificity matters. DoD topics describe precise technical requirements — sometimes down to specific operating frequencies or platform integration constraints. If your technology matches, you are competing against a small pool. NSF and NIH topics are broad, which attracts a wider and more variable applicant pool.

Third, resubmission dynamics. NIH explicitly allows and even encourages resubmissions, and revised proposals have historically won at higher rates than first submissions. If you have strong reviewer feedback from a previous cycle, a resubmission to NIH can significantly improve your odds. Most DoD topics, by contrast, are one-cycle opportunities — the topic either appears in a solicitation or it doesn't.

First-Time Applicants vs. Repeat Winners

The data on first-time versus experienced SBIR applicants tells a consistent story: repeat applicants win at higher rates. Across all agencies, companies that have previously received at least one SBIR award win subsequent Phase I proposals at roughly 1.3 to 1.5 times the rate of first-time applicants.

This does not mean the system is rigged. Experienced applicants have learned what reviewers look for, how to scope Phase I feasibility work, and how to write commercialization plans that satisfy evaluation criteria. They also tend to have stronger preliminary data and existing customer relationships.

The 2026 reauthorization introduced provisions aimed at leveling this playing field, including proposal caps that limit how many active SBIR awards a single company can hold simultaneously. The SBA's implementation guidance, expected by mid-2026, will determine exactly how these caps work — but the intent is to create more room for new entrants.

For first-time applicants, the strategic takeaway is to target agencies and topics where your technology is a strong match, not where success rates are highest in aggregate. A 15% hit rate at NIH for a proposal that directly addresses a published clinical need is a better bet than a 25% rate at an agency where your technology is a stretch fit.

Using the Data to Build a Submission Strategy

Smart SBIR applicants do not submit to one agency and hope. They build a portfolio strategy across two or three agencies over 12-18 months. Here is how to use success rate data in that planning.

Start by identifying every agency that funds research in your technology area. Browse SBIR opportunities and past awards on SBIR.gov to see which agencies have funded work similar to yours. For each agency, note the Phase I award size (NIH: $275,000; DoD: typically $50,000-$250,000 depending on component; DOE: $200,000-$250,000; NSF: up to $275,000) and the typical timeline from submission to award notification.

Then assess competitive density. How many proposals does this agency typically receive for topics in your area? Agencies publish award statistics, and SBIR.gov's award search lets you see how many awards were made per topic in recent cycles. A topic that funded 2 out of 8 proposals is a very different opportunity than one that funded 12 out of 80.

Factor in your team's experience. If this is your first SBIR submission, consider starting with an agency that offers pre-submission support — NSF's Project Pitch gives you early feedback, and many DoD program managers will discuss topic fit before you invest weeks in a proposal.

Review the complete SBIR application guide to understand the proposal components each agency requires, and explore SBIR-specific resources to see how other startups have navigated the process.

Related SBIR reading:

• SBIR Grant Guide 2026
• SBIR vs. STTR: Which Program to Choose
• 10 Mistakes That Get SBIR Proposals Rejected

The agencies with the highest success rates are not always your best target — the agencies where your technology solves a problem they have explicitly asked about are. Granted can help you match your capabilities to the right solicitations and build proposals calibrated to each agency's evaluation priorities.