Research Proposal Sample PDFs: A Directory of Real Funded Examples

A funded proposal PDF in your discipline, paired with the reviewers' summary statement, is worth more than a stack of generic templates. The most pedagogically valuable items go a step further: the same proposal in its rejected form and its funded resubmission, with reviewer feedback for both. Most of these PDFs exist in public archives, but they're scattered across funder pages, individual researchers' Figshare uploads, and community-maintained lists that move every few years. This page is the directory of where they actually live, organized by funder type. For a longer essay on how to read these effectively, see successful grant application examples; for the structural anatomy reviewers expect, see anatomy of a winning grant proposal.

§ IUS Federal Agencies

NIH is the only US federal agency that systematically publishes funded full proposals — and only through specific institutes (NIAID is the deepest archive; NCI, NHGRI, NIA, and NIDCD each maintain their own). NIA's K99/R00 and SBIR pages are particularly valuable because they release rejected and funded versions of the same proposal. NSF, DOE, and the rest publish abstracts and outcomes; PI-shared copies fill the gap. NEH, NEA, and IMLS are the humanities/arts/cultural-heritage trio.

§ IIEuropean Funders (ERC, MSCA, Horizon Europe, UKRI)

The ERC does not publish funded proposals; the practice of releasing them comes from individual awardees. The Austrian FFG keeps the most useful index of who has done so. MSCA Individual Fellowships are better represented than full ERC grants because more fellows have deposited their submissions. Pair this with the ERC Starting Grant playbook.

§ IIIFoundations & Charities

Foundations rarely publish full applicant narratives. The Wellcome Open Research Fund 2018 and 2019 releases are the clearest exceptions — entire competition cohorts with funded and unfunded applications side by side. Wenner-Gren's Proposal Collection is the best-curated funder-side anthropology archive on the web. Open Philanthropy's grant write-ups are written by the funder rather than the applicant, but they show what a thoughtful donor considers a credible case.

§ IVCrowdsourced & Personal Archives

Where official archives stop, peer-shared collections start. These are the most uneven sources — and often the most useful, because they include unusual mechanisms (DOD CDMRP, NIH F-series, niche foundations) the funders themselves don't publish, and several contain the rare rejected→funded pairs.

§ VFellowships & Predoctoral Awards

Fellowship samples are abundant because awardees share them more freely than PIs share R01s. NSF GRFP dominates by volume; ACLS and NEH cover humanities; Hertz, NDSEG, PD Soros, and Gates Cambridge show up regularly via individual awardees.

§ VIOther Regional Funders (abstracts only)

Japan, Korea, China, Taiwan, and Latin America generally don't publish full funded narratives. The two abstract-level databases below are the best public access for those regions. Australia and Canada have institutional "Successful Grants Libraries" but all known instances are gated behind university logins.

¶How to read these without copying them

The point of reading a funded proposal is not to mimic its sentences; it's to internalize the moves the writer makes. A useful pass looks like this:

• Read the summary statement first. If a critique is attached, you'll see exactly which choices the panel rewarded — and which it tolerated. That's higher signal than the narrative alone.
• Mark the structural moves, not the prose. Where does the gap statement land? How long is the preliminary data section relative to the rest? Where is innovation argued? These patterns transfer; sentences don't.
• Compare two proposals in the same mechanism. A single example is anecdote; two from the same call show what's structural versus stylistic.
• Use rejected→funded pairs first when you can find them. The NIA K99/R00 page, NIA SBIR page, Sylvain Deville's 2009/2010 ERC pair, John Bunce's NSF Cultural Anthropology pair, UNC TIBBS's F31 pair, and the Wellcome 2018/2019 cohort releases are the clearest worked examples of what changes between near-miss and funded.

¶What this directory deliberately doesn't include

Sites that aggregate "grant proposal templates" without provenance are excluded — if a sample isn't tied to a real awarded grant number, an institution, or a funder release, it's a stylistic exercise dressed up as evidence. Same for AI-generated example libraries: they show what a generator outputs, not what reviewers fund.

Several agencies and foundations also have to be excluded because they simply don't publish full funded narratives, despite often being asked about. Naming them here so you don't waste time searching:

• NIH institutes other than NIAID, NCI, NHGRI, NIA, NIDCD: NHLBI, NEI, NIMH, NIDA, NIDDK, NIAMS, NICHD, NIGMS, NINDS, NIAAA, and NIEHS do not maintain public sample-application libraries and generally redirect applicants to the institutes that do.
• DOD CDMRP, USDA NIFA, EPA STAR, NASA ROSES, NIST, ARPA-E, DARPA, IARPA: award abstracts only, or no public release. CDMRP and ED have FOIA reading rooms — slow and redacted, but possible.
• Continental European national funders: DFG, ANR, SNSF, NWO, FWO, FNRS, FCT, Academy of Finland, Research Council of Norway, Swedish Research Council, Vinnova, MIUR/PRIN, FWF, AEI, NCN, GAČR — none publish funded narratives. LaTeX templates are the most you'll find publicly. Use FFG's compilation and individual researcher Figshare deposits instead.
• UK research councils (BBSRC, MRC, NERC, EPSRC, ESRC, AHRC, STFC): no funder-direct release of case-for-support PDFs. Cancer Research UK publishes interview-style applicant case studies, not proposals.
• Canada and Australia: NHMRC, ARC, CIHR, NSERC, SSHRC don't publish narratives. University libraries (Melbourne, Western Sydney, Queen's, McGill, UBC, SFU) maintain successful grants collections, but all known instances require staff/student login.
• Asia and Latin America: JSPS, JST, AMED, NRF, IBS, NSFC, Academia Sinica, FAPESP, CNPq, CAPES, CONACYT, ANID, CONICET — abstracts only at best.
• Major US foundations (HHMI, Bill & Melinda Gates beyond TB MAC, Mellon, Ford, MacArthur, Sloan, Hewlett, Carnegie, Pew, Templeton, Burroughs Wellcome, Damon Runyon, Searle, Beckman, Keck, Simons, Chan Zuckerberg Initiative, Allen Institute, Schmidt Sciences): no funder-direct release. Some grantees post their own; check Open Grants and individual lab pages.

For specifics on each major mechanism's structure and review criteria, see the dedicated guides on NIH R01 Specific Aims, ERC Starting Grant, NSF directorates, and grant budgets.