R21 (NIH Exploratory/Developmental Research Grant)

An R21 is an NIH research-grant activity code — the Exploratory/Developmental Research Grant Award — that funds early-stage, high-risk research testing a novel idea, an untested model system, or an emerging tool or technology, before that work is ready for the scale and rigor NIH expects of an R01. R21 support is capped NIH-wide at $275,000 in total direct costs over a project period of no more than two years, with no more than $200,000 in direct costs in any single year. Most R21 Funding Opportunity Announcements (FOAs) explicitly state that preliminary data are not required, which is the mechanism’s defining administrative feature relative to R01. An R21 can be awarded on its own, or serve as the exploratory first phase of a milestone-gated, two-phase award paired with an R33 (the combined “R21/R33” mechanism).

The R21/R33 phased structure

Where a Funding Opportunity Announcement uses the combined R21/R33 model, an applicant submits a single application describing both phases. The R21 phase — up to two years, within the $275,000/$200,000 caps above — supports feasibility work, proof-of-concept testing, or generation of the initial data needed to justify expanded development. If the awarding NIH Institute or Center (IC) confirms that the predefined, quantifiable go/no-go milestones set out in the original application were met, the award transitions administratively into the R33 development phase — without a new competing application. Sponsored-programs offices should treat a combined R21/R33 as a single grant number with two internal budget and reporting phases, not two separate awards; the go/no-go decision point is an internal NIH administrative action, not a new Notice of Award requiring fresh proposal routing. A newer variant, R61/R33, uses an R61 activity code in the exploratory role R21 plays here, mainly in NIH’s translational and early-phase clinical-trial FOAs.

Budget, duration, and preliminary-data policy

The $275,000 total-direct-cost cap and $200,000-per-year sublimit apply NIH-wide to R21 unless a specific FOA sets a different figure (some ICs run more restrictive R21 programs). Because the per-year cap sits below the $250,000 threshold NIH uses for its Modular Budget format, most R21 applications are prepared as modular budgets rather than itemized detailed budgets. The distinguishing programmatic feature is the preliminary-data policy: R21 FOAs are written for ideas that are, by design, too early-stage to have generated the kind of supporting data an R01 study section expects — an application over-supplying preliminary data can read, to reviewers, as better suited to an R01. “Not required” does not mean reviewers ignore feasibility; R21 applications are still expected to make a credible case that the proposed exploratory work is achievable within the two-year, capped-budget period.

How R21 differs from R01

Both are R-series research grants carrying the same investigator-independence model, but they serve different points in a research idea’s lifecycle. An R01 supports a discrete, already-de-risked project chosen and defined by the investigator, typically built on substantial preliminary data and reviewed for renewal via a competing Type 2 application. An R21 exists to fund the exploratory step before that point — testing whether an idea, model, or tool works at all — within a fixed two-year, $275,000 ceiling, and it has no renewal mechanism: when the project period ends, the PI must pursue a new award (a fresh R21, a follow-on R01 built on the R21’s findings, or, where the governing FOA offers one, transition into a paired R33 phase) rather than file a competing renewal.

Standard due dates and resubmission

R21 applications (along with R03, R33, R21/R33, R34, R36, and several cooperative-agreement codes) follow the standard due-date cycle of February 16, June 16, and October 16 for new applications, and March 16, July 16, and November 16 for resubmissions, renewals, and revisions (5:00 PM local time of the applicant organization). NIH permits one resubmission (A1) of a given new or revision application, which must be received within 37 months of the original application’s receipt date or NIH treats it as a new submission rather than a resubmission. See the NIH standard due dates entry for the full activity-code table.

Typical use cases

Sponsored-programs offices and investigators typically reach for R21 when: a research idea is genuinely novel or high-risk and would not currently compete well for an R01; the goal is proof-of-concept or feasibility testing rather than a full research program; or a translational or clinical pipeline needs a formal, milestone-reviewed bridge (via R21/R33) between an exploratory finding and committed development funding, without the 12-18 month gap and administrative restart of a fully separate competing application. Whether a given IC’s payline or award strategy makes R21 competitive in a specific fiscal year is IC- and cycle-specific — see NIH payline for how that threshold works, and the CASRAI NIH grants overview guide for how R21 fits among NIH’s broader set of research, career-development, and multi-project mechanisms such as the P01 program project grant.

References

• National Institutes of Health, Grants & Funding, “Exploratory/Developmental Research Grant Award (R21)” and “Exploratory/Developmental Grants Phase II (R33)” activity-code pages, grants.nih.gov (primary source; grants.nih.gov returned HTTP 403 on direct automated fetch this session, so budget/cap and preliminary-data figures here are corroborated via multiple independent secondary sources rather than a fresh direct read — treat as REPORTED tier pending a manual primary-source check).
• National Institutes of Health, Office of Extramural Research, “Activity Codes, Organization Codes, and Definitions Used in Extramural Programs” — R-series classification and modular-budget threshold.
• NIH standard due dates and resubmission-policy notices (NOT-OD-12-128), corroborated via NIH Center for Scientific Review guidance and multiple institutional research-office publications of the same figures.