NSF Turns SBIR/STTR Back On With $250M, a July 27 Deadline, and a New $40M Instrumentation Lane: What the Two-Solicitation Architecture Means for Deep-Tech Founders

The National Science Foundation's Small Business Innovation Research and Small Business Technology Transfer programs are back in operation. On June 2, 2026, NSF reopened its Project Pitch portal — the mandatory pre-application step that has gated every NSF SBIR submission since 2019 — and posted two distinct solicitations covering a combined $250 million in fiscal year 2026 awards. The first full-proposal deadline is July 27, 2026, with secondary windows on November 4, 2026 and March 4, 2027.

The headline detail most coverage has skimmed past: NSF did not publish one solicitation. It published two. NSF 26-510 is the general deep-tech solicitation that follows the program's traditional structure. NSF 26-511 is a brand-new $40 million instrumentation pilot that runs a parallel review track for companies building scientific equipment. Founders must pick the correct lane before they pitch — submissions to the wrong track will not be administratively reassigned, and a misfiled pitch wastes the one-shot review NSF provides per concept per cycle.

For deep-tech founders, the relaunch ends a roughly eight-month period in which NSF SBIR functioned as a closed door while the program waited on Congressional reauthorization, a new merit review framework, and the foreign-risk management requirements that came out of the late-2025 reauthorization. The pent-up demand is real. The two-solicitation architecture is the most significant structural change NSF has made to the program since it imposed the Project Pitch requirement six years ago, and it changes which companies should be at the front of the queue.

What $250M actually buys

NSF's $250 million FY26 envelope is allocated across the standard SBIR award ladder, plus the new instrumentation lane and a small invitation-only Strategic Breakthrough tier carried over from prior cycles. The award sizes that most companies will plan around:

• Phase I: Up to $305,000 over 6–18 months. Feasibility-stage work to demonstrate that a high-risk technical concept can plausibly clear the engineering or scientific hurdles required for commercial development.
• Phase II: Up to $1.25 million over roughly 24 months. The serious R&D commitment, available only to companies that have completed Phase I (or are submitted as Fast-Track combined Phase I/II proposals).
• Fast-Track: Up to $400,000 in Phase I plus $1.155 million in Phase II for companies that can demonstrate both technical readiness and commercial pull at the time of pitch.
• Strategic Breakthrough: Up to $30 million per company, invitation-only, for a small number of Phase II awardees whose technology has demonstrated outsized commercial trajectory.

Historical Phase I funding rates at NSF SBIR have hovered between 10% and 20%, depending on the technical area and the cycle. Founders should plan around that distribution rather than treating the $305K Phase I ceiling as the realistic expected value of a single pitch.

The Project Pitch chokepoint

Every NSF SBIR submission begins with a Project Pitch — a short, structured concept document that NSF reviews internally and either invites to full proposal or declines. The pitch step is not advisory. It is a binding gate, and the pitch-to-invitation conversion rate has been the under-discussed metric in NSF SBIR for years.

NSF's stated turnaround on pitches is one to two months. Founders aiming at the July 27 full-proposal deadline therefore need to have a pitch submitted by late June or early July to receive an invitation in time to write a competitive proposal. Companies that wait until July to pitch will be writing for the November 4 cycle, not the July 27 cycle.

The strategic implication: the July 27 deadline is functionally a June pitch deadline for any company that does not already have a pending or recently invited pitch. The portal opened June 2; the time arbitrage is small but real.

NSF 26-511: the instrumentation lane

The new instrumentation pilot is the most structurally important change in this cycle and the most easily misunderstood. NSF 26-511 targets companies developing next-generation scientific equipment in categories that NSF explicitly describes as poorly served by traditional venture capital — mass spectrometers, cryogenic electron microscope accessories, microfluidic platforms, quantum sensing hardware, and a long tail of related instrumentation categories where the customer base is the research community itself.

The $40 million pilot reflects a long-running NSF concern that scientific instrumentation development has been crowded out of the SBIR program by software, biotech, and energy companies whose commercial trajectories are easier to articulate in standard pitch templates. The instrumentation lane runs a parallel review with reviewers selected for fluency in research-instrument markets, where revenue scales differently, sales cycles are longer, and the relevant customer is often a single grant-funded principal investigator rather than an enterprise buyer.

For founders deciding which lane to file under, the practical test is who buys the device. If the customer is a research lab paying with grant funds, NSF 26-511 is the correct track. If the customer is an industrial buyer or a commercial product team, NSF 26-510 is the correct track even when the technology has scientific origins. The split is about commercial pathway, not about technical sophistication. Cryo-EM accessories sold to academic structural biology labs go to 26-511. Cryo-EM-derived machine vision sold to semiconductor inspection vendors goes to 26-510.

Eligibility: the lines that disqualify before review

NSF SBIR eligibility rules are strict and they are the single most common reason companies get desk-rejected before scientific review begins. The lines that matter:

• Company size: Fewer than 500 employees, including all affiliates and majority-owned subsidiaries. The affiliate rule trips up companies that took strategic investment from a corporate parent.
• U.S. ownership: At least 50% owned by U.S. citizens or permanent residents. Companies majority-owned by multiple venture capital firms, private equity firms, or hedge funds are excluded. This is NSF's distinctive rule — it does not exist at DoD or NIH — and it is the rule that most often surprises founders who have raised institutional capital.
• Work location: All funded R&D must be performed in the United States.
• Principal Investigator employment: PI must be employed at least 20 hours per week by the company and committed to at least one calendar month (173 hours) per six months of project duration, with at least 51% of the PI's total employment with the company.
• STTR-specific: STTR requires a research institution subaward partner, with the work split mandated at 40% to the small business and 30% to the research institution at minimum, leaving 30% flexible.

Founders who have raised a priced venture round should run the ownership math before pitching. A standard Series A that takes the institutional cap table over 50% disqualifies the company from NSF SBIR even when no single investor crosses the threshold, because NSF aggregates institutional ownership in its eligibility test.

Reauthorization adds foreign-risk and performance gates

The late-2025 SBIR reauthorization layered two new compliance requirements on top of NSF's standard eligibility test. Foreign risk management requires companies to disclose certain foreign relationships and affiliations and certify the absence of disqualifying ties. Performance benchmarks apply to multi-award winners — companies with significant prior SBIR history must show commercialization progress to remain eligible for new Phase I awards.

For first-time NSF SBIR applicants, the performance benchmark rule does not bite. For repeat awardees, it does, and the practical implication is that companies sitting on stale Phase II awards without follow-on revenue will face questions in this cycle that they did not face in prior cycles.

Strategy for July 27 versus the November cycle

The honest framing for most founders evaluating NSF SBIR right now is whether to push for July 27 or aim at November 4.

Push for July 27 if: the company has a pending or recently invited pitch; the PI is identified and on payroll; ownership math is clean; the technical narrative is already written down in pitch form; and the company can commit four to six weeks of senior engineering and writing time in late June and July.

Aim at November 4 if: the pitch has not been submitted yet and the technical narrative is still being refined; the company is finishing a fundraise that may change the ownership math; the PI hire is in progress; or the company is choosing between the 26-510 and 26-511 tracks and needs reviewer-pattern data from the July cycle before committing.

The November cycle is not a consolation prize. It is the cycle most companies should target, and the July 27 deadline is the cycle for companies that have been waiting on NSF to reopen for eight months and already have the materials in a drawer.

For broader context on how NSF SBIR fits into the FY26 federal R&D landscape, see Granted's coverage of NSF's $8.75B FY26 budget and the parallel NIH SBIR/STTR omnibus reset running on a September 8 deadline.