NSF Funds Nine of These a Year. The Other Door Into the Same Program Has No Competition at All.

The NSF Research Experiences for Teachers (RET) in Engineering and Computer Science program takes proposals for RET Sites on October 14, 2026 — the second Wednesday in October, and annually thereafter. The governing solicitation is NSF 24-503.

The program's arithmetic is unforgiving. NSF estimates approximately nine Site awards per year against roughly $5.8 million in annual funding, with each Site capped at $600,000 over up to three years and $200,000 in any single year. Nine awards. Nationally. Across every engineering, engineering technology, computer science, and information science department in the United States.

Most institutions reading that should not submit a Site proposal this year. Almost all of them should be using the program anyway, through a channel that involves no competition and that a large majority of eligible PIs never touch.

The supplement is the real opportunity

RET operates on two tracks, and the solicitation title says so: Supplements and Sites.

A RET Site is a standalone center-scale award — a structured summer research program hosting a cohort of K-14 educators, with its own recruitment, mentoring architecture, curriculum translation, and academic-year follow-up. It is a full proposal, reviewed competitively, and it will beat out roughly nine in however many are submitted.

A RET Supplement is a request for additional funds on an existing NSF ENG or CISE award, at up to $15,000 per participating teacher or community college faculty member for one year, including up to $2,000 for classroom materials and equipment developed during the experience. It is not a competition against a national field. It is a request to the program officer who already manages your award, on research that NSF has already decided to fund.

The asymmetry is the point. If you hold an active NSF engineering or computer science award, the marginal cost of bringing a high school physics teacher or a community college engineering technology instructor into your lab for a summer is a conversation with your program officer and a modest supplemental request against work already underway. If you do not hold one, a Site proposal is a nine-slot lottery.

There is no published annual deadline for supplement requests the way there is for Sites, which is exactly why they get neglected — nothing on the calendar forces the decision. The practical guidance is to raise it with your cognizant program officer well before the summer you intend to host, because the funds must be requested, approved, and in place before a teacher can be paid. October is not too early for a summer 2027 conversation. March is late.

Who can actually submit a Site proposal

The Site eligibility rules eliminate a substantial share of the people who would naturally write this proposal, and they do it on the cover page rather than in review.

Submitting organizations are institutions of higher education — two-year and four-year, explicitly including community colleges. That inclusion is real and underexploited; a community college with an engineering technology program and a nearby research university partner is a credible applicant, not a courtesy mention.

The PI must be full-time, tenured or tenure-track faculty in engineering, engineering technology, or computer and information science, whose primary responsibilities are research and teaching — not administrative.

Read that last clause twice. The people who typically assemble K-12 outreach proposals at a research university are associate deans, center directors, directors of diversity and outreach programs, and research professors on soft money. Under NSF 24-503, none of them can serve as PI on a RET Site. The proposal must be led by a working faculty researcher who will actually mentor in a lab. NSF is enforcing the program's theory of change through the eligibility rule: teachers are supposed to do research, in a research group, with a researcher — not attend a professionally administered summer institute.

An individual may be listed as PI or co-PI on no more than one RET Site proposal per annual deadline. In a large department where several faculty want in, that constraint has to be resolved internally before the submission, not discovered at the portal.

What a Site has to contain

The required program elements are specific enough to function as a budget test:

• 8 to 10 or more K-14 educators engaged in an engineering, computer science, or information science research project.
• A minimum six-week summer duration.
• A well-defined research focus — not a survey of the department's activities.
• High-quality mentoring strategies.
• Academic-year follow-up activities connecting the summer experience back to classrooms.

Industry partners are encouraged but not required — worth noting, because applicants routinely assume an industry letter is a gate. It is a strengthener, not a requirement, and a weak or generic industry letter is worse than none.

Now put the elements against the cap. At $200,000 per year for ten educators, the program has roughly $20,000 per participant per year to cover stipends, materials, mentoring time, coordination, evaluation, academic-year follow-up, and indirect costs. Indirect alone consumes a meaningful fraction at most research institutions. A proposal that promises twelve teachers, a rich evaluation contract, and an ambitious dissemination plan on a $200,000 annual line is signaling to reviewers that the applicant has not costed the work.

One more budget rule that catches people: voluntary committed cost sharing is prohibited. School districts and industry partners will offer in-kind contributions, and the instinct is to put a dollar value on them to demonstrate commitment. Do not. Describe the partnership substantively — released time, facilities access, teacher recruitment channels, curriculum adoption commitments — without quantifying it as committed cost share.

Why this deadline still exists

There is a structural oddity worth flagging for anyone tracking NSF this year. On August 17, 2026, NSF replaced twelve of its core disciplinary solicitations with new notices of funding opportunity carrying more than $1.5 billion, and made them "Proposals Accepted Anytime" — abolishing the fixed submission calendar that most research groups had organized around for decades. We covered that reset and its consequences in the foundational research NOFO analysis.

RET did not move. It keeps a hard annual deadline, it is still governed by NSF 24-503 — a solicitation issued in 2024 and running unamended into its third cycle — and it sits outside the consolidation entirely.

Two implications follow. First, the deadline-abolition mental model does not transfer: the programs that lost their deadlines are the core disciplinary research programs, and the education, workforce, and infrastructure programs around them mostly did not. If you have recalibrated your lab's submission calendar around "anytime," RET is one of the items that will fall off it.

Second, because the solicitation predates two years of subsequent NSF policy movement, applicants must reconcile NSF 24-503's text against the current PAPPG for proposal preparation requirements, and should not assume that language about project descriptions, evaluation plans, or participant support costs written in early 2024 reflects current guidance. Where the solicitation and the current PAPPG diverge on mechanics, the PAPPG governs. Where the solicitation states program-specific requirements — the six-week minimum, the cohort size, the PI eligibility rule — those hold.

The case for applying anyway

Nine awards is a discouraging number, but the field is smaller than the odds suggest, for a reason specific to this program: the PI eligibility rule removes the institutions whose outreach capacity is concentrated in administrative offices, and the six-week, cohort-scale, academic-year-follow-up requirements remove the departments that wanted to run a two-week summer camp. What remains is a field of working research groups with genuine district partnerships — a real but not enormous population.

The applicants with the strongest position are:

• Departments with an existing teacher pipeline — an established relationship with a district or a community college partner, evidenced by prior teachers who have already worked in the labs. NSF is buying durable relationships between universities, community colleges, school districts, and industry, and prior evidence of durability is the most persuasive thing a proposal can carry.
• Community colleges as lead or co-lead, where the K-14 framing is native rather than aspirational. A Site that serves community college engineering technology faculty as participants is addressing a workforce layer with very little other federal support.
• Groups that have already run RET Supplements. This is the sequencing that works: use supplements on an existing award to build the mentoring practice and the district relationship over two or three summers, then submit the Site proposal with a track record instead of a plan.

That last path is the one to internalize if the October 14 deadline is too close to meet well. A rushed Site proposal against a nine-award field is a poor use of four weeks. A supplement conversation with your program officer this month, followed by two teachers in the lab next summer and a Site proposal in October 2027 with real outcomes to report, is a materially better expected value — and it puts teachers in labs a full year sooner either way.

Program contact for Site questions is RET_ENG@nsf.gov. For supplements, the only contact that matters is the program officer on your existing award.