On August 3, Secret Houston reported on a RestaurantData.com study finding that Houston led the United States and Canada in restaurant closures over the first half of 2026. The figures, as published:
Market | Estimated closures, H1 2026 |
|---|---|
Houston (city) | 119 — first in US and Canada |
New York City | 117 |
Chicago | 82 |
San Antonio | 70 |
Dallas | 42 |
Harris County | 172 — first of any county in US or Canada |
Houston–Sugar Land–Baytown MSA | 239 — fourth nationally |
Texas (state) | 1,039 — first of any state or province |
Closures are half a number. The other half is what opened.
The net
Every business in Texas that sells liquor by the drink files gross receipts with the state every month. Count the distinct venues filing above $0 in Harris County each December and you get the balance of both flows — everything that opened, minus everything that stopped.

Houston has posted a net gain every single year since 2020. Five consecutive years, +854 venues. The county has 3,412 venues filing today — the most in the twenty years of record, and 26% more than the 2,708 that filed the month before COVID.
The report's own window
The study covers January through June. Run the filings over the identical six months and the comparison gets sharper.
January → June | Net change in venues filing |
|---|---|
2021 | +222 |
2022 | +84 |
2023 | +72 |
2024 | +85 |
2025 | +59 |
2026 | +88 |
The six months in which a national study named Harris County the closure capital of North America produced the largest first-half gain in venues of any normal year since the COVID rebound. Not one major Texas market shrank over the same window:
January 2026 → June 2026 | Change |
|---|---|
Harris County | +88 |
Houston–Sugar Land–Baytown MSA | +149 |
Dallas County | +92 |
Bexar County | +42 |
Travis County | +38 |
Tarrant County | +31 |
Texas, all counties | +767 |
Why both can be true, and why one is more useful
The two datasets aren't looking at the same businesses. The report's own description says its closures skew quick-service, sandwich concepts, small-footprint locations, and over half were chain-affiliated. Almost none of those hold a mixed-beverage permit. A sandwich franchise closing was never in the state's liquor file to begin with.
But that's a difference in scope, not a defense of the method. Two problems remain with a ranking built on closures alone.
It ranks size, not distress. The report says this outright — it "indicates total closure volume rather than localized failure rates" — and then the number gets published as a league table. Houston is one of the largest restaurant markets on the continent. Houston at 119 against New York at 117 is a two-restaurant gap between two enormous markets, presented as first place.
It has no denominator and no opposite. A city that closed 119 and opened 200 is growing. A city that closed 119 and opened 40 is in trouble. The same closure count describes both, which means the closure count on its own describes neither.
The turnaround that isn't
There is a squeeze in this data, and it deserves more attention than the closure ranking. Compare first halves like for like — same months, every year:
First half | Venues filing | Per venue, Jan–Jun | Per venue, Jan–May |
|---|---|---|---|
2022 | 2,881 | $53,253 | $53,767 |
2023 | 3,006 | $54,586 | $54,869 |
2024 | 3,170 | $52,073 | $52,321 |
2025 | 3,278 | $50,027 | $50,139 |
2026 | 3,362 | $50,404 | $49,932 |

Read the Jan–Jun column and 2026 looks like the bottom of the cycle: $50,404 per venue, up 0.8%, the first gain since 2023 after two years of real decline. That number does not survive contact with the calendar.
June 2026 was the eighth-largest month in the nineteen years of record. Six of Houston's seven World Cup matches were played at NRG Stadium inside it. June came in at $52,722 per venue against a January–May average of $49,932, and accounted for 17.7% of first-half receipts — against 16.1% to 16.5% in every prior year on record.
Take June out and run the same five months across years. January–May 2026: $49,932 per venue, down 0.4% on the same five months of 2025. Not a bottom — a third consecutive annual decline, and 9.0% below the January–May 2023 peak.
The compression did not stop. It was papered over for one month by the largest sporting event on earth, and that month landed inside the window everyone is quoting. Rooms did close — plenty of them, and the report is probably right that there were a lot. More opened. And the ones still filing are, with the World Cup carved out, still losing ground to each other.
Which is the same lesson, pointed back at us: a six-month window with an anomaly in it will tell you whatever the anomaly wants.
One caveat
A venue missing from a recent month hasn't necessarily closed — it may have filed late, sold, or hit a processing lag. So we report the net, which is reliable, and we don't build a closure count out of a file that hasn't settled. Month over month, five years running, the direction of that net has not changed once.
Data: TABC Mixed Beverage Gross Receipts, data.texas.gov, dataset naix-2893, January 2007 – June 2026. Venue counts are distinct businesses by name and ZIP filing receipts above $0, with multiple permits at one address counted once. This covers businesses holding a mixed-beverage permit, not all restaurants. Recent months remain subject to revision as the state posts late filings.
Closure figures are as reported by Secret Houston on August 3, 2026, citing RestaurantData.com. We have not seen that study's methodology and reproduce its published figures as stated.
The publisher holds an ownership interest in one or more venues appearing in these rankings. All venues are ranked by filed receipts, without exception.


