Fast-food chains in New York City, by city inspection records

What stands out

Four in ten first inspections miss an A, citywide. 40.2% of initial inspections in the city’s file scored 14 points or more. The letter in a restaurant’s window is a different measure: for those restaurants it is settled at a second visit, after they have had the chance to fix what was found. This page is about the first visit, when nobody knew the inspector was coming.

The lowest group cannot be put in order. Starbucks (10.1%), Taco Bell (12.5%), Chipotle (14.4%) and McDonald's (14.4%) have the four lowest measured rates. Every one of their ranges overlaps every other, so the records do not support calling any of them lowest. Taco Bell’s range is the widest of the four, 8.5% to 18.1%, because it has the fewest inspections: 184.

Starbucks is the clearest case of the menu showing up in the numbers. A counter that mostly pours drinks and reheats pastry has fewer ways to lose points than a kitchen handling raw chicken. Its low figure is real, and it is not evidence that its stores are better run than a fried-chicken chain’s.

Dunkin' is the best-measured chain in the city. With 1,738 initial inspections, more than any other chain here, its range is only 16.2% to 19.8%. When a chain has that many restaurants, its figure says more about the chain’s system and less about a handful of operators.

Papa John's is level with the city. At 39.1% against 40.2% for all restaurants, with a range of 31.2% to 47.6%, we cannot tell it apart from an ordinary New York restaurant. That is a statement about 133 inspections at its franchised locations here, and nothing wider.

Many national chains are too small here to chart. Several chains that are everywhere in other states have only a few dozen restaurants in the five boroughs, so 16 chains on our list fall under the minimum and appear without a bar.

Analysis written September 2026. The figures in it are read from the current records, so they move when the records do.

How inspections work in New York City

The city’s Health Department inspects every restaurant on a cycle. Each violation carries points, and the points add up to a score: lower is better. A restaurant that scores 0 to 13 on its initial inspection earns an A and is next inspected in 11 to 13 months.

A restaurant that scores 14 or more is not graded on the spot. It is reinspected at least seven days later, and the reinspection score sets the grade: A for 0 to 13, B for 14 to 27, C for 28 or more. Its next cycle also starts sooner, in three to seven months depending on the score. So a restaurant with a poor first inspection is visited more often, and shows up more often in any count of inspections, this one included.

Two limits come from the city’s file itself. It holds only restaurants that are still in business on the day it is published, so places that closed are gone from it, along with their inspections. And it keeps roughly three years of history per restaurant, which is why 2023 is thinner than the years after it.

How these figures are made

1. Source. DOHMH New York City Restaurant Inspection Results, published by the NYC Department of Health and Mental Hygiene.
2. What counts. Only initial cycle inspections; re-inspections, complaint and pre-permit visits are left out. Inspectors go back to restaurants that already had a problem, so counting return visits would stack the numbers.
3. The figure. For each chain, the number of those inspections that scored 14 or more points, so did not earn an A on the first visit, divided by all of them. “All restaurants” is the same sum over everything in the file, chains included.
4. The range. A 95% Wilson score interval. If we could rerun the same period many times, the true rate would land inside a range built this way about 19 times in 20. Fewer inspections means a wider range.
5. The minimum. A chain needs 100 counted inspections to be charted. Below that it is listed with its count and no bar.
6. Matching names to chains. Restaurant names are matched to chains with a published table of name patterns (version 2026-09-20.3). A location recorded under two chain names, such as a combined KFC and Taco Bell, is counted for neither. If a match is wrong, write to hello@fastfoodfact.com and it will be corrected.
7. Reuse. New York City's Open Data Law (Administrative Code section 23-502) makes public data sets available without restrictions on their use, and lets the city require a reuser to name the source and version and describe any modifications, which this site does.

The full method, including the formula for the range, is on the methodology page.

Scoring, grades and the inspection cycle: NYC Health, Inspection Cycle and Letter Grading and How We Score and Grade. What the file includes and leaves out: the dataset’s own description on NYC Open Data. We count inspections typed “Cycle Inspection / Initial Inspection” and group the file’s violation rows into one record per restaurant per date before counting.

What this does not show

• Management quality on its own. Chains differ in what they handle. A coffee counter deals with far less raw food than a kitchen that fries chicken, so part of any gap between chains is the menu.
• Anything about one restaurant. A figure describes that chain’s locations in New York City over this period. It is not a statement about the brand elsewhere and not a warning about the one near you.
• Every restaurant equally. Inspections are scheduled by risk and by past results, so a restaurant with problems is visited more often and appears more often in these counts. That holds for chains and for all restaurants alike.
• How New York City compares with anywhere else. Each place counts a different event under different rules. These figures only make sense next to other figures from New York City.

Informational only. These are public records, summarized; they are not a safety rating of any restaurant and not professional advice. Last reviewed: September 2026.