The $68 Million NSF Program That Pays Colleges to Turn Scientists Into Teachers — and Just Waived Its Cost-Share Requirement

The United States has a math and science teacher shortage that no amount of exhortation has fixed. High-need school districts — the ones serving the highest concentrations of low-income students — routinely cannot staff chemistry, physics, and upper-level mathematics classrooms with qualified teachers. The people who could fill those roles, STEM majors and working scientists and engineers, can almost always earn more doing something else, and the cost and friction of getting certified is enough to end most of the conversations before they start.

The National Science Foundation's Robert Noyce Teacher Scholarship Program is the federal government's most direct attempt to change that math. It makes roughly $68 million available to colleges and universities to build the pipeline that moves STEM talent into K-12 classrooms in high-need districts — through scholarships, stipends, and teaching fellowships — with individual awards ranging up to about $3.25 million. The full-proposal deadline falls on the last Tuesday in August each year; for 2026 that is August 25.

And this cycle carries a change that materially lowers the bar to entry: NSF has waived the cost-sharing requirement for Track 2 and Track 3 proposals submitted through August 2026, and prohibited voluntary committed cost sharing on any new Noyce proposal in that window. For institutions that have looked at Noyce before and balked at the matching burden, that single line rewrites the calculus. Here is how the program works and how to compete for it.

What Noyce funds, and who it is really for

Noyce is not a scholarship a student applies for directly. It is a grant to an institution — a college, university, or consortium — that then runs a program to recruit, prepare, and support future STEM teachers. The applicant is the department or school of education (usually in partnership with STEM disciplinary faculty and a high-need local education agency), and the money flows through the institution to the scholars and fellows it recruits.

The program's stated mission is precise: to address the critical need for recruiting, preparing, and retaining highly effective elementary and secondary mathematics and science teachers in high-need Local Education Agencies (LEAs). That phrase "high-need" is load-bearing. A high-need LEA is defined by concentration of poverty, and Noyce scholars incur a service obligation to teach in exactly those districts. This is not a program for staffing well-resourced suburban schools; it is explicitly targeted at the districts that struggle most to attract STEM educators.

The program runs across four tracks, plus a capacity-building option:

• Track 1 — Robert Noyce Teacher Scholarships and Stipends. Scholarships for STEM undergraduates and stipends for STEM professionals to become K-12 teachers. This is the classic Noyce mechanism.
• Track 2 — NSF Teaching Fellowships. Fellowships for STEM professionals and recent graduates pursuing teacher certification and a master's degree, with sustained support as they enter and stay in high-need classrooms.
• Track 3 — NSF Master Teaching Fellowships. Support for experienced, certified STEM teachers to become master teachers and leaders in high-need districts.
• Track 4 — Noyce Research. Funds research on the recruitment, preparation, and retention of STEM teachers — the evidence base for the rest of the program.
• Capacity Building. For institutions that want to develop a full proposal in any track but need to build the partnerships and infrastructure first.

That structure means Noyce serves institutions at very different stages. A university with no existing teacher-prep pipeline can enter through Capacity Building; one with a mature program can go straight for a multi-million-dollar Track 1 or Track 2 award.

Why the cost-share waiver is the story this cycle

Buried in the mechanics is the change that should get more attention than it will: Track 2 and Track 3 cost sharing is waived for proposals submitted through August 2026, and no voluntary committed cost sharing is allowable on any new Noyce proposal in that window.

To understand why this matters, you have to understand what cost sharing does to an applicant pool. When a federal program requires an institution to match a portion of the award with its own funds, it filters the field toward institutions with money to spare. A well-endowed research university can commit institutional funds to a Teaching Fellowship match without breaking stride. A regional public university or a minority-serving institution — often the ones with the strongest ties to the high-need districts Noyce targets — frequently cannot, and self-selects out before writing a word.

Waiving the requirement for Tracks 2 and 3 removes that filter for exactly the cycle in front of us. Institutions that previously concluded they could not afford to compete for a Teaching Fellowship or Master Teaching Fellowship award should re-run the analysis, because the financial barrier that produced the "no" is gone for this deadline. The prohibition on voluntary committed cost sharing reinforces the point: NSF is signaling that it does not want proposals competing on who can pledge the biggest institutional match. It wants them competing on the quality of the program design.

Waivers like this are typically temporary. Treating this cycle as a distinct opportunity, rather than assuming the terms will persist, is the correct posture — especially given the broader budget pressure across federal research.

What a competitive Noyce proposal actually requires

Noyce is a partnership program, and the proposals that win are the ones that demonstrate a genuine, functioning collaboration rather than a hopeful sketch of one. Three elements separate fundable proposals from the rest.

A real high-need LEA partnership. The single most common weakness in Noyce proposals is a district partnership that exists on paper only. Reviewers want evidence that the high-need LEA is a co-designer, not a letterhead — that district leaders have shaped the program, committed to hosting and mentoring scholars, and have a plan to actually employ them. The service obligation only works if there are real jobs in real high-need schools at the end of the pipeline, and the proposal has to make that concrete.

Integration of STEM and education faculty. Noyce sits at the seam between a university's STEM departments and its school of education, and the strongest programs are genuinely co-owned. A proposal driven entirely by education faculty, with disciplinary scientists nowhere to be found, signals a program that will produce teachers with weak content preparation — precisely what Noyce is trying to avoid. Reviewers look for disciplinary depth in the mathematics and science preparation, not just pedagogical training.

A retention plan, not just a recruitment plan. The teacher shortage is as much a retention problem as a recruitment one; STEM teachers leave high-need districts at punishing rates. Proposals that stop at "we will recruit and certify N scholars" miss half the program's intent. The competitive ones describe induction support, mentoring, communities of practice, and the concrete mechanisms that keep new teachers in high-need classrooms past the brutal first few years.

Where Noyce fits in the 2026 funding landscape

Noyce's August 25 deadline sits at the tail end of an unusually dense NSF education-and-workforce stretch. The Research Training Groups and CREST-family deadlines cluster in early-to-mid August, and NSF's broader FY2026 portfolio — anchored by an $8.75 billion budget and plans for roughly 10,000 new awards — is being spent against a policy backdrop that has made federal research funding more contested than it has been in years.

That context cuts two ways for Noyce applicants. On one hand, STEM workforce development enjoys unusually durable bipartisan support; the argument that the country needs more qualified math and science teachers in poor districts is about as politically safe as federal education spending gets. On the other, the government-wide grant framework faces an October 2026 overhaul that adds new layers of political review to discretionary awards and scrutinizes how programs frame their target populations. Noyce's high-need-district framing is grounded in poverty concentration rather than demographic category, which positions it comparatively well, but applicants should write toward the program's core statutory purpose — effective STEM teachers in high-poverty schools — and let that mission carry the proposal.

For an institution weighing whether to commit the effort, the combination this cycle is favorable: a large, mission-safe program, awards up to $3.25 million, four tracks that accommodate institutions at every stage of readiness, and — for the tracks most institutions found hardest to afford — a waived cost-sharing requirement that will not necessarily be there next year. The full proposal is due August 25, 2026. For universities with a genuine high-need district partner and the disciplinary depth to prepare teachers who actually know their subjects, this is one of the clearest-return opportunities on the summer NSF calendar.