Short answer
Start six months out. Confirm the program fits a current announcement and your institution's limits, assemble trainee and faculty data first, write the program plan around measurable aims, build in rigor, career development and evaluation, and finish a week before the NIH date so your sponsored programs office can submit.
Should we apply at all?
Apply only when your program clearly fits a current announcement, your institution meets its eligibility rules, and you can show trainee outcomes or a credible plan to produce them. Before anyone drafts text, answer four questions:
- Which announcement? For NIGMS in 2026 that usually means the predoctoral T32 (moving to the Parent T32 and eight program areas from January 25, 2027), BURT for undergraduates, TURTLE for Tribal entities, or a parent R25.
- Are we eligible? BURT, for example, is limited to institutions under $50 million a year in NIH research project grants and allows one application per organization. The eligibility quick-check flags the common rules.
- Is there a conflict? NIGMS allows one predoctoral T32 per program area per institution. Your sponsored programs office will know about limited-submission rules.
- Who leads? Program directors need protected time. Writing and then running a training grant is not a side task.
What should we assemble first?
Data, before prose. Reviewers judge a training program largely by its record: who applied, who was admitted, how long students took, where they went next and what the faculty have published with trainees. Pull these early:
- applicant, admissions and enrollment numbers for the last several years
- trainee outcomes, including degree completion, time to degree and first positions
- faculty research support and training records
- current and past trainee publications
If those numbers live in five offices, the timeline builder schedules the data pull at about 16 weeks out for good reason.
What goes in the program plan?
A program plan should read as a set of specific training aims with activities and measures for each, not a catalog of everything the department does. Typical parts:
| Part | What reviewers look for |
|---|---|
| Training aims | Three to five concrete outcomes the program commits to |
| Program structure | How trainees are selected, appointed, advised and moved through the program |
| Courses and skills | Required coursework, including rigor and reproducibility, and professional skills |
| Mentoring | How mentors are chosen, trained and held accountable |
| Career development | Individual development plans, exposure to careers, networking |
| Inclusive environment | Recruitment, climate and retention practices, described concretely |
| Evaluation | How you'll know each aim is being met |
Since NIH's 2015 notice on rigor and transparency, institutional training grants have been expected to describe formal instruction in rigorous experimental design. Our rigor training guide covers ways to do that well.
How do we write the evaluation section?
Tie every training aim to at least one measure you already collect or can start collecting next semester. A short logic model helps reviewers see the link between activities and outcomes. Our program evaluation guide walks through one.
What schedule actually works?
| Weeks before NIH date | Milestone |
|---|---|
| 26 | Go/no-go decision; limited-submission check |
| 16 | Data pull starts |
| 12 | Program plan and evaluation plan drafted |
| 8 | Letters of institutional commitment and support requested |
| 6 | Data tables and biosketches done |
| 4 | Full draft out for internal review |
| 1 | Final to sponsored programs office |
The timeline builder turns that into dated milestones for your chosen due date, moves NIH deadlines that fall on weekends or federal holidays, and exports to your calendar.
Questions people ask
How long does a training grant application take?
Plan on about six months for a new application and three to four for a renewal with good records. The data tables alone can take several weeks if trainee outcomes aren't already tracked in one place.
Do NIGMS reviewers care about the number of faculty?
They care about whether the faculty can train well, which shows up as funding, publications with trainees, past trainee outcomes and mentor training. A smaller group with strong records reads better than a long list of names with thin training histories.
When should we apply to renew an NIGMS predoctoral T32?
NIGMS recommends applying a year to a year and a half before the current award ends, because predoctoral T32s are funded once a year with a July 1 start.