The Small Business Grants Study
We publish a database of 413 small business funding programmes. This page reports what is in it. Every programme was read against one written rule, not the label on the listing. Then we checked whether a business can still apply.

Key Takeaways
- 1227 of 413 published programmes are grants a business can apply for and keep.
- 2181 are something else, loans, tax credits, equity deals, advisory services, or money only a city, university or nonprofit can apply for, and 5 fit no rule cleanly.
- 3Of the grants, 76 are expired, paused or dead linked, and 35 could not be confirmed either way.
- 4116 programmes are both a grant and confirmed open, which is 28.1 percent of the database.
- 5Every figure here is derived from live data, so it moves when the data moves. Last verified September 6, 2026.
227 of the 413 published programmes (55.0 percent) are grants a business can apply for and keep. Another 181 are something else. Loans, tax credits, equity deals, advisory services, or money that only a city, a university or a nonprofit can apply for. The remaining 5 fit no rule cleanly, and we count them in the total rather than dropping them.
Then there is the second cut. Of the 227 grants, 76 are expired, paused or point at a dead link. Another 35 we could not confirm either way. That leaves 116 programmes that are both a grant and confirmed open.
So out of 413 entries, 116 are a live grant a business can actually apply for. That is our own data, audited against ourselves.
Published programmes
413
Are grants
227
55.0 percent of all published
Grants confirmed open
116
28.1 percent of all published
Every published programme, grouped by what kind of money it is. We did not audit anyone else's list. We audited ours, because it is the one we can prove.
| Kind of funding | Programmes | Share |
|---|---|---|
| Grant | 227 | 55.0 percent |
| Not business facing | 106 | 25.7 percent |
| Loan | 25 | 6.1 percent |
| Technical assistance | 23 | 5.6 percent |
| Tax credit | 18 | 4.4 percent |
| Equity investment | 9 | 2.2 percent |
| Not classified | 5 | 1.2 percent |
| All published programmes | 413 | 100 percent |
The status split across the 227 programmes that are grants. Dead means one of three things. The programme expired, it was paused, or the application link no longer resolves. Unconfirmed means we could not establish either way at the last check, and we count those separately rather than assuming they are open. It also holds rounds announced for a window that has not opened yet. Those are a confirmed not open rather than an unknown, and the label is rougher than we would like.
| Status | Grants | Share of grants |
|---|---|---|
| Confirmed open | 116 | 51.1 percent |
| Could not confirm | 35 | 15.4 percent |
| Expired, paused or dead link | 76 | 33.5 percent |
| All grants | 227 | 100 percent |
Our rule is one question. Can a small business apply for this and receive money it does not pay back? If yes it is a grant. If the money is repaid it is a loan, including loan guarantees. If the programme takes a stake it is an equity investment. If it reduces a tax bill rather than paying money out it is a tax credit. If it buys advice, training or consulting rather than handing over funds it is technical assistance.
The last category surprises people. A programme is not business facing when the applicant is a city, a county, a tribe, a university, a nonprofit or an economic development body, even where a small business benefits in the end. The test is who signs the form, not who benefits. It is also the biggest non grant category we hold, 106 programmes.
Two calls here move the number, so we state them plainly. A forgivable loan counts as a grant only when the debt is wiped on terms the applicant can meet alone. If someone decides later whether to wipe it, we call it a loan. Calling a loan a grant is the error that saddles a founder with debt they did not expect. So we take the safer side. A matching grant is still a grant. You put in your own money. The awarded part is never paid back.
Some programmes do not fit any rule. We leave those unclassified rather than guess, and we count them in the total rather than dropping them. Dropping them would flatter the percentage. There are 5 of them, and they fall into three groups. Programmes that bundle a loan, an equity fund and an advisory service into one listing, where no single label is honest. Programmes whose own page has gone and cannot be read. And programmes where the source contradicts itself, naming businesses as eligible in one place and barring them in another.
Anyone checking this study against one of our state pages will find a bigger open number there, and the difference is scope rather than standard. A state page counts every kind of funding it holds for that state, loans and tax credits and advisory programmes included. This page counts grants and nothing else. The two were never measuring the same population.
The tests themselves are now the same on both surfaces. Expired, paused and dead linked strike a programme off here and dim it on the state page. So does a note that argues with its own status. So does a round announced for a window that has not opened. The last of those differences closed on September 3, 2026, when a programme nobody had confirmed recently stopped counting as open on the state pages too. One case could still part the counts. A programme with no status at all would be held back here and shown there.
You can check any state yourself. The state grant pages list every programme we hold for that state with its own status, and our grants guide covers how to apply for the ones that are open.
The Runway Number is our other open data page. It works out what six months without a salary costs in each state, living costs, health cover and the first year business fee added together. Same method, our own data.
Every figure on this page is computed from the live table each time the page loads, so it moves when the data moves. Nothing is typed in by hand. The stamp on this reading is September 6, 2026. A verifier rechecks programme status continuously, and the classification improves as programmes are re read, which means the numbers here have moved before and will move again.
Nothing on this page is a recommendation and nothing here earns us a commission. There are no partner links on it by design.
Frequently Asked Questions
Sources
Every figure on this page is computed from our own grants database at the moment the page is served, using the same rules that decide what the state grant pages show. Copy last reviewed September 4, 2026.
About the Author

Business Formation Researcher
Eliot leads StartupOwl's state data research. He maintains the site's fifty state records of LLC filing fees, annual costs, processing times, and small business grant programs, checking each figure against the state office that publishes it and logging the date it was verified. Based in New Jersey, he has spent his working life in and around small businesses, and he writes for the founder who wants the real number rather than the advertised one. Every figure under his byline traces to a named source, and when a state proves a number wrong, the correction is published, not buried.