Letter of Inquiry vs. Letter of Intent: Two Different Documents That Share the Same Abbreviation

A client asks you to "write the LOI." Depending on who the funder is, that instruction means one of two very different jobs. It might mean drafting a one-to-two-page document that a program officer will actually score, deciding whether your client gets to submit a full proposal at all. Or it might mean filling out a short administrative form that no reviewer will ever read and that, for at least one major federal funder, doesn't even get accepted anymore. Both documents get called "LOI." Treating them as interchangeable is the mistake worth avoiding.

This guide separates the two uses of the abbreviation, explains a real, recent change on the federal side, and lays out what actually goes into the version that matters most for the freelance and independent grant writers this site is written for: the foundation letter of inquiry.

The federal "letter of intent": an administrative courtesy, and one major funder just dropped it

Some federal funding announcements have historically asked applicants to file a short letter of intent before the full application deadline — typically listing the project title, the principal investigator, participating institutions, and a brief abstract. According to NIH's own December 2025 policy notice, this kind of LOI "was never required, binding, nor considered during the peer review process." Its only real job was administrative: helping agency staff estimate how many applications were coming and recruit the right reviewers ahead of time (National Institutes of Health, Notice NOT-OD-26-019).

That job just went away at NIH. Effective December 3, 2025, per Notice NOT-OD-26-019, "NIH will no longer request or accept Letters of Intent (LOIs) as part of the application process." The notice explains the change as part of a broader effort to reduce applicant burden now that peer review has been centralized through NIH's Center for Scientific Review, which removed the need for the workload-estimating purpose the LOI used to serve. It's also explicit about what happens to LOIs submitted anyway: they "will not be acknowledged nor reviewed" by the relevant institute or center (National Institutes of Health, Notice NOT-OD-26-019).

Two caveats worth knowing if you write proposals across different federal agencies. First, this change is NIH-specific — grant-research site OpenGrants notes that other federal agencies and specific programs, including some at the Department of Education, the National Science Foundation, and the Department of Energy, still use preliminary submissions such as letters of intent or concept papers to manage applicant volume, and whether one is required varies notice by notice (OpenGrants, "Letter of Intent for Grants: The Rules Just Changed"). The only reliable way to know is to read the specific funding announcement's application and submission section rather than relying on general advice, including this article. Second, OpenGrants also flags that the NIH change doesn't touch a different document that happens to share a similar name: a subrecipient or consortium commitment letter, which documents a partner institution's role in a project budget and is unaffected by the LOI policy change.

The foundation "letter of inquiry": the version that can end an application before it starts

Private and family foundations use the same three letters for a document that does the opposite kind of work. GrantStation, a nonprofit-sector grants resource, defines it plainly: "a letter of inquiry is a brief, formal document that nonprofit organizations send to a funder to introduce a proposed project and request permission to submit a full proposal," adding that funders also call it a letter of interest, a letter of intent, or a concept paper "depending on the grantor" (GrantStation, "What Is a Letter of Inquiry for Grants?"). That last point is worth sitting with: some foundations genuinely use "letter of intent" as their own name for this competitive screening document, which is exactly why checking the specific funder's own instructions matters more than memorizing a universal rule for what each label means.

How many foundations actually gate on this step is a case where two credible data sources tell slightly different stories, worth reporting honestly rather than flattening into one number. Candid's 2025 Foundation Giving Forecast Survey asked 522 private and community foundations directly whether they accept unsolicited proposals; 55% said yes (56% of private foundations, 48% of community foundations) (Candid, "How often do foundations accept unsolicited requests for funds?"). But the same Candid analysis also looked at IRS Form 990-PF filings — where private foundations must disclose whether they only fund preselected organizations and don't accept unsolicited requests — for 112,733 private foundations that filed for fiscal year 2023, and found 71% checked that box, meaning a narrower reading puts the share open to unsolicited requests at closer to 29% (Candid, same source). Candid itself notes the two figures aren't measuring quite the same thing: the survey sample skews toward larger, more active grantmakers, and respondents may interpret "unsolicited proposals" more loosely than the IRS checkbox language. The practical takeaway either way: a meaningful share of foundations use some kind of inquiry or gatekeeping step before a full proposal, and it's never safe to assume a given foundation accepts proposals cold without checking their own guidelines first.

What "gatekeeping" looks like in practice, from a funder's own account: Peter Marks, executive director of the L.B. Research and Education Foundation (a $35 million family foundation), described his foundation's process on the National Center for Family Philanthropy's blog: a 500-word emailed concept paper, roughly 70 of them per quarter from first-time applicants, of which "about 10% lead to requesting a full proposal" — and once a full proposal is requested, better than even odds of being funded (Peter Marks, "Declining Applications and Inquiries: How Funders Can Do Better," National Center for Family Philanthropy). That's one foundation's numbers, not an industry average, but it illustrates exactly why OpenGrants calls the foundation inquiry "the one that can sink you": most of the selection happens at the inquiry stage, not the full-proposal stage, so treating it as a low-effort formality is a real, costly mistake (OpenGrants, "Letter of Intent for Grants: The Rules Just Changed").

What actually goes into a foundation letter of inquiry

GrantStation lays out five components most funders expect, in roughly this order: a brief introduction that establishes credibility (mission, history, relevant track record); a project description and statement of need; an explicit connection to the funder's own stated priorities; budget information (the total project cost and the amount requested, without necessarily a full line-item breakdown); and contact information with a professional close (GrantStation, "What Is a Letter of Inquiry for Grants?"). The same source's list of common mistakes is worth reading as a checklist in reverse: letters that run too long or unfocused, letters generic enough to send to any funder (a sign of skipped research), letters that skip the specific requested amount, and letters with typos or missing required attachments.

Length is the detail most likely to trip up a freelance writer working across several funders at once, because it varies more than general advice suggests. GrantStation puts most foundation LOIs at one to two pages. But OpenGrants points to real counterexamples that both come from the funders' own guidelines: grantmaking-software firm Foundant Technologies recommends a three-page limit with budget detail as a separate attachment, while the W.T. Grant Foundation's 2026 research grant application guide asks for a full five-page letter of inquiry (OpenGrants, "Letter of Intent for Grants: The Rules Just Changed," citing Foundant Technologies and the W.T. Grant Foundation directly). There's no safe default length to memorize — the only reliable number is whatever the specific funder's own guidelines state that cycle.

A practical rule for freelance grant writers juggling both kinds

OpenGrants frames the decision as a short sequence, and it holds up well for freelance work spanning federal and foundation clients in the same week: identify the funder type first. If it's NIH, there's currently no LOI step to write at all. If it's another federal agency, open that specific funding announcement and check its submission section for any required preliminary document — don't assume based on what a different agency, or an older guide, used to require. If it's a private or family foundation, assume there may be a real inquiry gate, and confirm the format, length, and deadline on the funder's own site before drafting anything (OpenGrants, "Letter of Intent for Grants: The Rules Just Changed"). Matching your effort to the actual stakes follows from the same logic: a required federal preliminary document deserves accuracy, since it's rarely scored; a foundation letter of inquiry deserves the same strategic care as a full proposal, since for a meaningful share of foundations, it effectively is the first round of one.

A short checklist before you send either one

• Have you confirmed, on the funder's own current guidelines or the specific funding announcement, whether an LOI is required, optional, or no longer accepted — rather than relying on a template or a memory of what this funder used to want?
• If it's a federal LOI, have you checked whether this specific agency and program still requests one, instead of assuming NIH's December 2025 change applies everywhere?
• If it's a foundation letter of inquiry, does it follow that funder's own stated length and format exactly, rather than a generic one-to-two-page default?
• Does the letter name a specific requested amount and tie it to the funder's own stated priorities in their own language, not just your organization's general mission?
• Have you avoided treating any LOI as low-stakes paperwork before checking which of the two documents it actually is?