A foundation’s website tells you what it hopes to fund. Its tax return tells you what it actually funded.
Those two things are often different, and the gap between them is where small nonprofits lose a lot of grant-writing hours. Imagine a development director who finds a foundation whose mission statement mentions “youth” and “education,” spends two weekends on a proposal, and hears nothing back. Later she learns the foundation has given to the same twelve organizations for six straight years.
The proposal wasn’t the problem. The prospect list was.
Every private foundation files a Form 990-PF with the IRS, and those returns are public. If you learn to read a few parts of that form, you can tell in about 20 minutes whether a foundation is worth a proposal, how much to ask for, and whether it ever funds anyone new. Here’s how I’d approach it.
Why the stated guidelines aren’t enough
Foundations matter to most fundraising plans. According to Giving USA, giving by foundations reached $117.15 billion in 2025, about 19% of all charitable giving in the United States. That’s a big pool, and it’s understandable that a small shop wants a piece of it.
But foundation guidelines are written broadly on purpose. “Strengthening communities” and “improving outcomes for children” could describe thousands of organizations. Guidelines tell you the boundaries. They rarely tell you the habits.
The 990-PF shows the habits. It lists every grant the foundation paid during the year, who received it, and how much. Read three years of those lists and you’ll see patterns no website mentions: the typical grant size, the geography, the kinds of organizations that get funded, and how often a new name shows up.
That last one is the most useful number in foundation research, and almost nobody calculates it.
Where to find the return
You don’t need a paid database to start. ProPublica’s free Nonprofit Explorer lets you look up a foundation by name and open its recent filings. It also offers full-text search across millions of electronically filed returns, including 990-PFs. Paid tools like Candid’s Foundation Directory make the work faster and add more history, so if you’re doing a lot of grant research, they can be worth the cost. For a first pass on a short list, the free filings are enough.
One caveat: a 990-PF is a look backward. Foundations file after their fiscal year ends, often with an extension, so the most recent return you can find may describe giving from a year or two ago. Board priorities and staff can change faster than that. Treat the return as strong evidence, not a guarantee.
The three numbers to pull from every 990-PF
You don’t need to read the return cover to cover. Look for three numbers and two checkboxes. Start with the numbers.
1. The giving budget. Private foundations are generally required to pay out a minimum amount each year, tied to a “minimum investment return” that the IRS defines as 5 percent of the value of the foundation’s investment assets (with some adjustments). Many foundations give more than the minimum, but the asset figure gives you a rough sense of scale. A foundation with $8 million in assets is probably giving something in the neighborhood of $400,000 a year. That’s useful context before you ask it for $150,000.
2. The typical grant, not the average grant. Scroll to the list of grants paid during the year in Part XIV. Don’t just divide the total by the number of grants. Averages get pulled up by a few big commitments.
Imagine that same $8 million foundation made 40 grants last year totaling $395,000. The average is about $9,875. But when you look closer, three grants account for $200,000 (a hospital, a university, and a community foundation). Twenty grants were $5,000, ten were $2,500, and seven were $10,000. The median grant is $5,000.
If you asked for $10,000 because that looked like the average, you’d be asking for twice what most grantees receive, from a foundation you’ve never met. A $5,000 request is far more likely to fit how the board actually decides.
3. The newcomer rate. This is the number I’d spend the most time on. Pull the grant lists from the last three returns. For the most recent year, count how many grantees did not appear on either of the two prior years’ lists.
If 4 of 40 grantees are new, that’s a 10% newcomer rate. In practical terms, this foundation has room for about four new organizations a year, and you’re competing with everyone else who found it. That doesn’t mean you shouldn’t apply. It means you should know the odds, and you should probably look for a way to get introduced rather than sending a cold proposal.
If the newcomer rate is zero across three years, you’re looking at a closed list. Your time is likely better spent elsewhere, at least for now.
The two checkboxes that save you weeks
Part XIV also includes a box a foundation checks if it “only makes contributions to preselected charitable organizations and does not accept unsolicited requests for funds.” If that box is checked, an unsolicited proposal probably won’t be read. That doesn’t make the foundation impossible forever. Relationships can open doors, and some foundations change their approach over time. But it should move the foundation off this quarter’s proposal list and onto a relationship list.
If the box isn’t checked, the same section usually tells you who to contact, what form an application should take, any deadlines, and any restrictions, such as geographic limits or fields the foundation won’t fund. Some small family foundations have no website at all, and this section is the only application guidance you’ll find. Read it before you write anything.
The second “checkbox” is one you create yourself: geography. Look at the addresses on the grant list. A foundation based in your state may give 90% of its money in one county. If you’re not in that county, its broad guidelines don’t matter much.
Use the grant lists in reverse
Here’s a tactic that works well for small shops. Instead of starting with foundations, start with organizations like yours.
Pick three or four peer organizations: similar mission, similar size, similar region. Search their names using ProPublica’s full-text search. Because foundations must list the grants they pay, the results can show you which foundations funded your peers. Those foundations have already shown they fund your kind of work at your kind of scale.
This also connects foundation research to something we’ve written about before: your best prospects usually already know you. Once you have a list of likely foundations, look at their trustees (also listed on the 990-PF) and share those names with your board. A board member who knows a trustee can turn a 10% newcomer rate into a real conversation.
A 20-minute worksheet
For each foundation on your list, write down:
- Total assets and an estimated annual giving budget
- Median grant size from the most recent year
- Newcomer rate across three years
- Whether the “preselected organizations only” box is checked
- Where most of the grant dollars go geographically
- Any peer organizations on the grant list
- Any trustee your board might know
Then sort. Foundations with an open application, a newcomer rate above zero, grants in your geography, and a median grant that fits your request go to the top. Closed foundations with a board connection go on a cultivation list. Everything else waits.
This is a less exciting way to spend an afternoon than writing a proposal. It’s also how you stop spending weekends on proposals that never had a realistic chance.
Where to start this week
Don’t research fifty foundations. Pick the five you were planning to apply to next. Pull three years of 990-PFs for each and calculate the median grant and the newcomer rate. There’s a good chance at least one drops off your list and at least one ask amount changes.
When you do get funded, the work shifts from finding the foundation to keeping it. That’s where a strong report matters, and we’ve covered how to write a grant report that sets up your renewal. If you want a broader look at researching individual donors too, the guide to donor prospect research is a good next read.
