ARPA-H's SBIR/STTR Deadline Is July 10 — But the Real Test Is a Two-Page Solution Summary, Not the Full Proposal.

There is a specific kind of health-technology company that has always fit awkwardly inside the federal funding system: too early for the FDA, too applied for the National Institutes of Health, too capital-intensive for a seed round that expects a demo in eighteen months. The company has a genuinely novel idea — a diagnostic that doesn't exist yet, a material with no predicate device, a platform that could work but has never been built — and it needs money that doesn't demand equity and doesn't demand that the science already be finished. That gap is precisely the one the Advanced Research Projects Agency for Health (ARPA-H) was created to fill, and its 2026 SBIR/STTR Broad Agency Announcement is the clearest on-ramp small businesses have into it.

The BAA puts seven high-ambition health topics on the table, backed by Phase I contracts of up to $600,000 and Phase II contracts of up to $3.5 million — all of it non-dilutive. But the structure of how you enter matters more than the dollar figures. The first gate is a Solution Summary due July 10, 2026 at 11:59 PM ET, and the overwhelming majority of the selection pressure lives right there, in a short document, long before anyone writes a full proposal. This is the deep dive on how the ARPA-H model actually works and how a small firm gets through the funnel.

Non-Dilutive, But Not NIH

The reflexive comparison for any health SBIR is the NIH program, and understanding why ARPA-H is not NIH is the single most useful thing a founder can internalize before applying.

NIH SBIR money flows through the institutes and, for all its rigor, rewards incremental, hypothesis-driven advancement — a well-powered study, a defensible aim, a logical next experiment. ARPA-H is built on the DARPA model instead: it funds non-incremental, transformative bets with an explicit commercialization pathway and an expectation of market impact. The agency sits inside the Department of Health and Human Services but behaves like a venture-style program office. It wants technical novelty and technical feasibility in the same breath — a thing no one has done, described convincingly enough that a skeptical program manager believes it can be done.

That difference shows up in the money itself. ARPA-H awards are issued as firm-fixed-price contracts with milestone-based quarterly payments, not grants. You are paid for hitting agreed technical milestones, not for effort. For a founder used to grant accounting, this is a real operational shift: the money is tied to deliverables, and a missed milestone is a missed payment. In exchange, you keep 100% of your equity and cap table — the defining advantage of non-dilutive funding, and the reason a $600,000 Phase I contract can be worth far more to a startup than the same amount of priced equity.

The Seven Topics

The 2026 BAA is topic-specific, not open-ended. You are not proposing whatever you like; you are answering one of seven defined problems, each reflecting an unmet clinical need ARPA-H has decided is worth a transformative swing:

1. An annual fertility biomarker test — a diagnostic giving women predictive insight into future fertility and reproductive planning.
2. Antimicrobial surgical adhesives — next-generation bioadhesives that are reversible and degradable.
3. A programmable microbial platform for toxin removal (UNI-PLAT) — a living, adaptive system for clearing toxins tied to chronic disease.
4. Non-invasive endometriosis therapy — a first-of-its-kind treatment addressing root causes rather than symptoms.
5. The ARPA-H Lineage topic — commercialization support for technologies that originated inside prior ARPA-H-funded efforts.
6. A rapid multi-system autoimmune diagnostic assay — fast, comprehensive testing for an area of profound diagnostic difficulty.
7. A virtual human brain model — a simulation platform for developing neurosurgical robotics.

Two structural signals are worth reading here. First, the heavy representation of women's health — fertility, endometriosis — reflects a deliberate agency priority toward historically under-funded areas where the science is genuinely open. Second, the Lineage topic is unusual and quietly valuable: it exists specifically to help small businesses commercialize technologies that came out of earlier ARPA-H programs. If your firm has any connection to prior ARPA-H work, that topic is a narrower, less-crowded lane than the open technical topics.

The Solution Summary Is the Whole Game

Here is the operational fact that decides who ever gets funded: ARPA-H uses a two-step application process, and the first step is the one that eliminates most applicants.

• Step 1 — Solution Summary, due July 10, 2026. A short document describing your concept, its novelty, its feasibility, and its fit to the topic.
• Step 2 — Technical Presentation / full proposal. Only firms whose Solution Summaries are encouraged proceed to pitch.

This inverts the effort curve that grant-writers are trained for. In a traditional SBIR, you pour weeks into a 15-page research plan and submit it cold. Here, the expensive full proposal is invitation-only, gated behind a two-to-few-page summary that a program manager reads quickly and judges on a single question: is this transformative and feasible, and does it answer the topic as written?

The implication for a small firm is stark. Do not spend July writing a full technical proposal for a topic you have not been encouraged on. Spend it making the Solution Summary undeniable: a crisp statement of the leap, an honest account of why it is hard, and the specific technical reason you — and few others — can make it. Vague ambition dies here. So does incremental thinking dressed up as novelty. The summary must read like the opening of a DARPA pitch, not the background section of an R01.

Who Can Actually Apply

SBIR/STTR eligibility is federal boilerplate, but it is disqualifying if you get it wrong. Applicants must be small business concerns:

• 500 or fewer employees, with majority U.S. ownership, performing the work in the United States.
• For SBIR, the principal investigator must be more than 50% employed by the small business.
• Work-share rules: SBIR Phase I requires at least 66% of the work be performed by the small business; Phase II requires at least 50%. STTR requires at least 40% from the small business and at least 30% from a partnering research institution.

The SBIR-versus-STTR choice is not cosmetic. If your technology depends on a university lab's people and facilities, STTR's 30% research-institution requirement formalizes that partnership and lets a PI keep a primary academic appointment. If your firm can carry the work in-house, SBIR's higher small-business work-share keeps control — and dollars — inside the company. Choose deliberately, because the work-share math shapes your budget and your subcontracts from day one.

The Support Most Founders Overlook

ARPA-H bundles something NIH generally doesn't: hands-on commercialization support. Successful firms get access to Entrepreneur-in-Residence guidance, regulatory and reimbursement assistance, customer-discovery support, and optional I-Corps participation. For a deep-science team that has never navigated an FDA pathway or a payer conversation, this scaffolding can be worth as much as the contract. It also tells you what ARPA-H is really buying: not a paper, but a product with a plausible route to patients and a market.

What To Do Before July 10

The BAA has been circulating in draft — meaning topics and mechanics were previewable before the portal formally opened for submission. Treat that as a gift, not a delay. The firms that win ARPA-H money are the ones who used the draft window to pressure-test which of the seven topics genuinely fits their technology and to draft a Solution Summary they could defend to a skeptical program manager.

Concretely, before the July 10 gate: confirm your SAM.gov registration is active (this alone can take weeks and locks out latecomers), pick the one topic you can win rather than the two you could plausibly enter, and write the summary around the leap — what is impossible today that your approach makes possible. If it survived that framing internally, it might survive ARPA-H. If it reads like a reasonable next experiment, it won't.

The money here is genuinely founder-friendly: no equity, milestone-paced, with a commercialization apparatus attached. But ARPA-H is buying transformation, and it decides who to believe in the space of a few pages due July 10. Everything else is downstream of getting that right.

Tracking non-dilutive health-tech funding? See Granted News for the latest ARPA-H and SBIR announcements, and search Granted for the full 2026 SBIR/STTR deadline calendar across every federal agency.