The Chicago Restaurant Index · August 2026
What Chicago is opening.
A monthly count of restaurant openings, taken from the City of Chicago’s own business-licence record and classified with the same model that scores an address. It covers openings only — the method section explains why nobody, including us, can honestly count closures on a monthly basis.
Openings by month
Every complete month in the record for the last two years. The most recent month is highlighted.
How long it takes to open
The licence record carries the day an application was filed as well as the day the licence began. Across 2,088 restaurants that opened between September 2023 and August 2026, the first food licence took a median of 41 days from filing to start. A quarter were through in 23 days. A quarter took more than 92.
Two things this is not. It is the licence step alone, not the whole build-out: we measured Chicago building permits running months ahead of the licence record, so the paperwork starts long before this clock does. And it counts applications that succeeded, because an application that never issued never appears in the record at all. Read it as the floor on the licensing wait for a restaurant that did open, not as the expected wait for one about to file.
Concepts gaining and losing ground
Share of all openings in the trailing 12 months against the 12 before, in percentage points. Share rather than count, because a category can open more places and still be losing ground. Categories with fewer than 20 openings across both windows are left out — at that size a swing is two restaurants, not a trend.
| Concept | Openings | Year before | Share change |
|---|---|---|---|
| Italian / Pizza | 38 | 21 | +2.3pp |
| Sandwich / Sub | 22 | 12 | +1.4pp |
| Bakery / Dessert | 42 | 32 | +1.2pp |
| Breakfast / Diner | 15 | 8 | +1.0pp |
| Fried Chicken | 33 | 26 | +0.9pp |
| Mexican | 51 | 65 | −2.4pp |
| Coffee / Cafe | 83 | 90 | −1.6pp |
| Japanese / Sushi | 13 | 22 | −1.4pp |
| Health / Juice | 11 | 15 | −0.7pp |
| Latin / Caribbean | 14 | 16 | −0.4pp |
Where they are opening
The same comparison by community area.
| Community area | Openings | Year before | Share change |
|---|---|---|---|
| HUMBOLDT PARK | 19 | 4 | +2.1pp |
| WICKER PARK | 24 | 12 | +1.7pp |
| LAKE VIEW | 38 | 29 | +1.1pp |
| AUSTIN | 18 | 11 | +0.9pp |
| WEST RIDGE | 19 | 13 | +0.8pp |
| ROGERS PARK | 5 | 16 | −1.7pp |
| LOWER WEST SIDE | 8 | 17 | −1.4pp |
| LINCOLN SQUARE | 13 | 19 | −1.0pp |
| BELMONT CRAGIN | 16 | 20 | −0.7pp |
| IRVING PARK | 11 | 15 | −0.7pp |
Watch one neighbourhood.
This page is the whole city. Pick the area you actually care about and we will email you when a new restaurant licence appears there, and stay quiet when none does. It covers openings only: a closure takes about three years to surface fully in the public record, and we will not send an alert we cannot stand behind.
Method, and why there are no closures here
An opening is a record: a business licence is issued, and it appears. A closure is an inference, and a slow one. A restaurant goes dark long before its licence lapses, and the health inspection that confirms the space is empty arrives later still.
We measured how slow, against a cohort of 4,292 Chicago restaurants whose outcomes are now fully known. One year after a closure, the public record reveals 18% of them. At two years, 33%. Detection is only complete at three years.
So a monthly tracker that put openings against closures would compare a near-complete number with one that is missing four fifths of its subject, and would report Chicago gaining restaurants every single month. It would be wrong, and it would be wrong in the flattering direction. That is why this page counts one thing and says so.
Survival is a separate question with a separate answer: the concept-fit model estimates it for a specific concept and operator, at the zip's resolution, validated out-of-time on openings it never saw. This index deliberately does not compute its own version — two implementations of the same number are two numbers waiting to disagree.
Counts exclude licences that are not restaurants, and consolidate the paper duplicates that appear when a second licence at one address is folded into the first. A month is published only once the record extends past its final day, so the most recent month is never a partial one. Figures are regenerated monthly from the same artefacts the product scores from.
Licence record through 2026-09-03. Index generated 2026-09-04. Free to cite with a link.
- Chicago licenses
- never
- Google enrichment
- never