NIH's New NextTech Program Funds Neurotech and Pain Devices Up to the First-in-Human Stage: $6M, 8 to 12 Awards, First Deadline February 1, 2027. Hypothesis-Driven Applications Will Not Be Reviewed.

On October 23, 2026, the National Institutes of Health posted RFA-EB-28-004, "NextTech: Translational Development of Next-Generation Diagnostic and Therapeutic Devices (R18 Clinical Trial Not Allowed)." The National Institute of Biomedical Imaging and Bioengineering leads it, and eight other institutes participate: NEI, NIA, NIAAA, NICHD, NIDCR, NIDA, NIMH, and NINDS. The money comes from the two NIH programs with set-aside funding for device work, the BRAIN Initiative and the HEAL (Helping to End Addiction Long-term) Initiative.

The terms:

• About $6,000,000 in fiscal year 2028 for about 8 to 12 awards (the Grants.gov synopsis says four; the notice's figure of 8 to 12 is the one to plan on)
• Budgets that "should rarely exceed $750,000 in direct costs per year"
• Up to three years per project
• No clinical trials
• Open to universities, nonprofits, small businesses, large companies, governments, and foreign organizations (but no foreign subawards; see below)
• Six due dates: February 1, 2027; September 17, 2027; May 26, 2028; February 1, 2029; September 19, 2029; and February 1, 2030

The first application cycle opens January 1, 2027 and closes February 1, 2027. Review takes place in July 2027, council in October 2027, and the earliest start date is December 2027, which falls in FY2028. Resubmissions are allowed.

The program's aim is narrow: to take a proof-of-concept neural or pain-treatment device to the point where it is ready for a first-in-human clinical trial, and no further.

What NextTech funds, and what it rejects

The scope begins at proof of concept and runs through "development, optimization and pre-clinical testing" up to readiness for clinical studies. The devices may be "non-invasive, minimally invasive, implantable, or a combination thereof." They may use "electrical, optical, magnetic, acoustic" or other signals, and they may be used with drugs or biologics only if those are already FDA-approved.

The notice lists what it will reject outright. These applications "will not be reviewed":

• Projects to develop or validate animal models, diagnostic procedures, biomarkers, rehabilitation strategies, small molecules, or biologics
• Projects "where the primary objective is to study scientific or clinical hypotheses," or to evaluate clinical safety or efficacy
• Projects "that will not conclude with a path to human use"
• Projects that "lack justification for the mechanism of the device's functionality"

The notice also says applications "should propose design-directed development of a new technology and not be hypothesis-driven." It excludes "basic research, or studies of disease mechanism or mechanism-of-action studies." The research strategy "must focus on technology development rather than mechanistic, hypothesis-driven research or biological aims."

Most academic neuroscience labs write R01-style applications with hypothesis-testing aims. That format fails here. A NextTech application reads more like an engineering development plan, with design inputs, target specifications, a risk register, and milestones.

BRAIN applicants must pick "only one CNS disorder or disease, even if the device proposed for development could be used for more than one." HEAL applicants must name the specific pain or opioid-use-disorder condition and population. They must also justify the neural target, such as a nerve fiber or spinal circuit, and show that the device will "focally interact with the anatomical target."

Engineering-grade specifications

The notice is explicit about how specific target specifications should be. It gives an example of what NIH wants:

"prototype survives 32 days under accelerated aging at 80 degrees C with less than 5% degradation in any one operational parameter"

It also says that measurable-but-open specifications ("we will measure how long it survives") and subjective ones ("our device will survive aging tests") are "significantly less desirable." Where a number is a best guess, applicants should say so and state their confidence.

Other requirements in the research strategy:

• A design-control framework: how requirements will be "developed, refined, and validated," how risks and timelines will be managed, and how resources will shift if work stalls
• A thorough risk analysis, ideally informed by a hazard analysis, covering component maturity, the biological basis of the treatment, timeline threats, and ethical limits on translation
• A justification for every custom component: "For each component to be developed within the project, applicants must justify why off-the-shelf components are inadequate"
• A comparison to the standard of care for the chosen condition and population
• An FDA milestone: the work should produce enough preclinical safety and effectiveness data "to inform at least one pre-submission meeting with the FDA"
• A needs assessment, in a separate attachment, or a plan to perform one early if the device is at a very early stage
• A Gantt chart attachment, which reviewers consider in the overall score

Letters, IP, and industry partners

The letters-of-support section is stricter than most NIH notices:

• Academic applicants need a letter from the technology transfer official who will manage the project's intellectual property.
• Multi-institution projects need a letter from each institution explaining how IP will be shared.
• Industry collaborators need a letter "from a high official within the private entity" covering device access, limits on studies, limits on sharing data "(including negative results)," and whether a license will be needed. Signed agreements are required before award.
• Letters "should not be generic." They must say what each partner has contributed so far and what it will provide.

Start these early. Tech-transfer offices and corporate legal teams do not turn around a negotiated letter in two weeks, and for a February 1 deadline, December holidays shorten the time available.

How reviewers will score it

The standard NIH criteria apply, with NOFO-specific questions added:

• Significance: how big an advantage the device offers over clinically available options, and how strong the rationale is for the biological target
• Investigators: "How sufficient is translational experience (e.g. regulatory, manufacturing, commercialization) represented to guide development of the device towards human use within five years?"
• Innovation: "To what extent do the applicants make effective use of existing resources, technologies, and capabilities to speed translation?" On this program, reusing proven technology counts in an application's favor under Innovation.
• Approach: the post-grant development plan, barriers to clinical adoption, stakeholder-informed design criteria, and whether "the regulatory plan [is] reasonable in terms of the path to market"

The Investigators question is where many academic teams will lose points. A lab of electrophysiologists and graduate students does not meet the "translational experience" bar. Add a co-investigator or named consultant with FDA regulatory, design-controls, or medical-device manufacturing experience, and use the multiple-PI option, which the notice encourages for teams that combine engineering, computational, and scientific expertise.

Eligibility details that trip people up

Foreign organizations can apply, but foreign subawards are not allowed. Under NIH's current policy, applications with foreign subawards or subcontracts "will be deemed noncompliant and will not be considered for funding." Unfunded international collaborators, foreign consultants, and equipment bought from foreign vendors are still allowed.

Companies are eligible. Small businesses and larger for-profits can apply directly. For a device startup, NextTech is an alternative to an SBIR Phase II. The budget is comparable, there is no small-business size test, and an academic co-PI can be included without STTR's allocation rules. In exchange, the company must accept BRAIN or HEAL data-sharing requirements and the program's coordination meetings, including the annual HEAL Investigators Meeting for HEAL awardees.

Indirect costs follow the existing rules. The notice says that, under the FY2026 Consolidated Appropriations Act, the indirect-cost provisions of 45 CFR 75 continue to apply to NIH awards. Budget at your negotiated rate.

Which deadline to target

With six due dates over three years, a weak first submission is a poor choice, because a resubmission costs a full cycle. Use these rules of thumb:

• February 1, 2027 suits teams that already have a working prototype, bench data, quantitative target specifications, and industry or tech-transfer letters in progress. You have about 14 weeks.
• September 17, 2027 suits teams that need a few months for a needs assessment, a regulatory consultant, or an industry agreement. A stronger first submission is worth more than an earlier slot.
• 2028 and later dates depend on BRAIN and HEAL appropriations. The FY2028 figure of $6 million is the only funding commitment so far.

The short version

RFA-EB-28-004 offers about $6 million for 8 to 12 three-year R18 awards to develop BRAIN and HEAL devices up to first-in-human readiness. Budgets should rarely exceed $750,000 in direct costs a year, nine NIH institutes participate, and there are six due dates starting February 1, 2027. It funds engineering development, not hypothesis testing. Applications need quantitative target specifications, a risk analysis, a regulatory path that includes an FDA pre-submission, letters from tech transfer and industry partners, and a team with real translational experience. Companies and foreign organizations can apply, but foreign subawards are not allowed.

Sources: the Grants.gov listing for RFA-EB-28-004 and its attached full announcement. For related NIH opportunities, see our analysis of PAR-27-003, the RM1 Collaborative Program Grant and current NIH funding opportunities.

If you are turning a neurotechnology or pain-device prototype into a NextTech application, Granted can help you convert aims into target specifications, structure the risk analysis, and draft a research strategy that reads as an engineering plan.