NHLBI's Catalyze Program Will Fund Your Device Prototype Without Making You Start a Company — But It Wants Someone Else's Money First

There is a specific kind of dead end in translational research that almost nobody funds out of.

You are an academic investigator. You have a device concept, a diagnostic assay, or a research tool that works in your hands and addresses a heart, lung, blood, or sleep condition. The basic-science funding mechanisms consider it too applied. The SBIR mechanisms require a small business you have not started and may not want to start. Industry wants a prototype you cannot build without funding. The gap between "interesting result" and "thing a company would license" is maybe eighteen months of unglamorous engineering, and almost no grant exists for exactly that stretch.

NHLBI Catalyze exists for exactly that stretch. Its Product Definition suite — RFA-HL-26-016 through RFA-HL-26-020 — has a receipt date of October 21, 2026, and its eligibility is deliberately wide: universities, nonprofits, small businesses, for-profit companies, tribal organizations, and faith-based entities. You do not need a company. You need a partner and a matching dollar.

The five doors

Catalyze Product Definition is not one opportunity. It is a grid, and picking the wrong cell is the most common way to waste a cycle.

Two axes organize the whole thing. Therapeutics versus devices/diagnostics/tools determines your budget cap — $400,000 per year for molecules, $300,000 for devices, which reflects real cost structure rather than a value judgment. Early versus late maturity determines your mechanism: phased R61/R33 if you still have discovery-flavored work ahead of you, direct R33 if your prototype exists and needs optimizing.

NHLBI intends to commit up to $4,466,000 in each fiscal year from 2026 through 2028, funding up to 8 new awards per year — roughly 24 awards across the three-year run. Spread across five RFAs, that is a small number. Treat it as such when you calibrate ambition.

One hard boundary applies across all five: clinical trials are not allowed. This is pre-human product definition. If your next step involves human subjects under a protocol, you are looking at the wrong family of announcements — NHLBI's early-phase clinical trial mechanisms are separate.

The requirement that actually decides your application

Read the RFAs quickly and the Accelerator Partner requirement looks like a letter of support. It is not. It is the review criterion the rest of your application is being measured against.

An Accelerator Partner is a third party with commercialization capability who has looked at your technology and agreed to be named. Depending on the specific announcement, that means an industry partner, a venture group, an incubator or accelerator organization, or a comparable entity with a credible path to moving a product forward. The program requires at least one for the direct-R33 announcements — HL-26-018 and HL-26-020 — at the time of application. For the phased announcements — HL-26-017 and HL-26-019 — you do not need one at submission, but you cannot transition from the R61 to the R33 phase without one. HL-26-016 does not require a partner at all.

Alongside it sits the cost match: NHLBI specifies a minimum 0.25:1 ratio of non-federal funds to federal direct costs. For every $100,000 in federal direct costs, $25,000 from somewhere else. On the direct-R33 announcements, a letter documenting that match is expected with the application; on the phased ones, it is a transition requirement, documented before R61 funds convert to R33 funds. Because NHLBI's own program materials describe the match as recommended in some places and as a transition requirement in others, confirm the exact obligation against the specific RFA text you are applying under — this is precisely the kind of detail that varies by announcement and sinks otherwise strong submissions.

Put the two together and the design becomes obvious. NHLBI is not evaluating whether your device is scientifically interesting. It is evaluating whether anyone with money and market judgment agrees it is worth developing. The partner and the match are the same test asked twice.

That test is why the October 21 date is tighter than it looks. Nine weeks is ample for writing. It is thin for negotiating a partnership agreement with an industry sponsor, getting a technology transfer office to bless a match commitment, and obtaining a signed letter that survives review — a chain with at least three parties, none of whom report to you.

How the R61/R33 gate really works

For HL-26-017 and HL-26-019, the phased structure is the whole appeal and the whole risk.

The R61 phase funds a defined milestone-driven stretch of work with no partner and no match required up front. That is a genuine gift: it lets an academic team do the derisking that makes a partner interested, funded by the award itself.

Transition to the R33 phase is not automatic and is not peer-reviewed in the usual sense. An NHLBI committee evaluates whether you have hit your milestones, whether you have produced a non-federal cost match letter demonstrating third-party investment, and whether you have identified an Accelerator Partner. Fail any leg and the money stops at the end of R61.

This structure rewards a specific kind of proposal writing that most academic PIs are bad at: milestones stated as pass/fail quantitative criteria, not as aims. "Characterize the sensor response" is an aim. "Demonstrate a signal-to-noise ratio of at least X across N samples spanning the physiological range, with coefficient of variation under Y percent" is a milestone. The committee reviewing your transition is looking for the second kind of sentence, and it will be reading it eighteen months from now without the benefit of your enthusiasm in the room. Applications that write aims into the milestone table are the ones that stall at the gate. The same discipline governs other NIH phased mechanisms — we walked through it in the context of the HEAL analgesic discovery R61/R33.

Catalyze versus SBIR: pick deliberately

The most common strategic error here is treating Catalyze and the NIH SBIR/STTR omnibus as interchangeable translational money. They are not.

SBIR requires a small business concern with the associated ownership, size, and employment tests, and it puts the PI's primary employment inside that company for Phase I. In exchange it offers larger budgets, a Phase II path, and eventually mechanisms like the Phase IIB Strategic Breakthrough Award.

Catalyze does not require a company at all. It lets the work stay in the university, under the university's indirect cost agreement, with the institution holding the IP — while still demanding the commercial validation that SBIR gets structurally by requiring a business. For an investigator who is not ready to leave the faculty, that is not a lesser option. It is the only option in the category.

The corollary: if you already have a company and a product, Catalyze's $300,000-per-year device cap will feel small next to SBIR. Use SBIR. Catalyze is designed for the pre-company moment, and it is efficient there specifically because it does not ask you to become a founder first.

What to do in the next nine weeks

Week one: pick your cell in the grid, and be honest about maturity. The single most common misfile is an early-stage concept submitted to a direct R33 — HL-26-018 or HL-26-020 — because the applicant did not want the R61 milestone gate. Reviewers see through this immediately, and the required partner letter is nearly impossible to obtain for technology that is not ready. If your prototype does not exist yet, the phased announcement is your announcement.

Weeks one through four: work the partner problem first. Start with your institution's technology transfer or commercialization office, then any company that has already contacted you about the technology, then regional accelerators and disease-focused venture groups. Ask early, ask specifically, and give them a one-page technical summary with a stated ask. A partner letter negotiated in week eight reads like one negotiated in week eight.

Weeks two through five: solve the match. Institutional commitment, foundation support, philanthropic gifts, and partner in-kind contributions are all plausible sources depending on the announcement's terms. This is a conversation with your dean or your CFO, and it moves at the speed of their calendar, not yours.

Weeks three through eight: write the milestone table before you write the research plan. Then write backward from it. Every Catalyze application is fundamentally a claim that a specified amount of money produces a specified, verifiable technical state by a specified date.

Throughout: talk to the program. Catalyze staff, reachable through the program's NHLBI mailbox, will tell you which announcement fits your maturity and whether your partner arrangement satisfies the requirement. That conversation costs an email and routinely saves a cycle.

The bottom line

Eight awards a year is not a generous program, and $300,000 a year is not a large budget. But Catalyze is one of the few places in the federal portfolio where an academic team can get paid to turn a working idea into a defined product without first becoming a business — and the October 21 date gives you three more cycles under announcements that run into late 2027 if you miss this one.

Miss it for the right reason, though. Miss it because your partner conversation needs another two months, not because you found out about the partner requirement in October.