New York State's Red Violations, and Why They Are Not a Grade

New York State does not score restaurants and does not grade them. Its inspectors mark each violation one of two colours, and the file records how many of each an inspection produced. In the Health Department’s own description of the field, a violation is of “type, critical Red (R) or non-critical Blue (B).”

This site counts whether a routine inspection recorded at least one red. Across 2023–2026, 25.1% of 115,627 inspections of food service establishments in the state’s file did.

What red covers

Red violations, as the county health departments that carry out these inspections describe them, “relate directly to factors that could lead to foodborne illness or injury,” while blue ones “do not directly cause foodborne illness or injury but could negatively affect the establishment’s operation.” The practical difference shows in what happens next: “It is generally expected that critical violations be corrected while the inspector is present at the facility.”

That last point matters for reading the figure. A red violation is not a finding that survived the inspection — most are fixed before the inspector leaves. It is a record that the condition existed when an inspector walked in unannounced, which is exactly what makes it worth counting across thousands of visits and nearly worthless as a judgement of one.

What the file leaves out

The state’s dataset is called statewide, and it is not. The department lists the exclusions itself: the file “excludes inspections conducted in New York City, Suffolk County and Erie County,” each of which runs its own inspection system. Between them those three cover a large share of the state’s population, so the place page and every figure here say “New York State outside New York City, Suffolk and Erie” rather than letting the name stand for the whole state.

Two more limits the department states: “Occasionally, remediation may not appear until the following month due to the timing of the updates,” and “Update frequencies and availability of historical inspection data may vary from county to county.” The second is the more serious one for a comparison like this, because it means the file’s coverage is not uniform across the counties inside it. A chain concentrated in a county that reports promptly is measured on more recent visits than one concentrated in a county that does not.

New York City is not a gap on this site — it has its own file, its own rules and its own page. But the two are not comparable, and nothing here adds them together. The city counts points on a first visit; the state counts whether a red violation was written down. Those are different events.

Why this is not a grading system

The Health Department says it directly, in the terms attached to the data:

“The inspection data contained in this dataset was not collected in a manner intended for use as a restaurant grading system, and should not be construed or interpreted as such. Any use of this data to develop a restaurant grading system is not supported or endorsed by the New York State Department of Health.”

That is a constraint this site takes literally. There is no per-restaurant lookup here and there will not be one: no address, no letter, no score for the place down the road. What the state’s file supports is the thing on this page — a count over thousands of inspections of how often a class of restaurant records a red violation, with its margin of error printed beside it. A grade is a verdict on one business. A band is a description of a population.

What the chain figures show

18 chains have enough counted inspections to be charted here, and 17 sit clearly below the 25.1% rate for all food service establishments in the file.

Note what that baseline covers. The state’s file is not restaurants alone; it is food service establishments, which is why this site never calls it “all restaurants” for New York State. Our counts exclude schools, camps, institutions and feeding programmes on both sides of the comparison, but the remaining population still runs from diners to banquet halls.

Dunkin' is the best-measured chain here, with 2,816 counted inspections putting it between 7.6% and 9.7%. And as everywhere on this site, part of the distance between a chain and the local rate is the menu: red violations concentrate in temperature, holding and raw handling, and a counter doing less of all three has fewer opportunities to record one.

Sources

• New York State Department of Health, Food Service Establishment Inspections: Beginning 2005 (Active) — the field definitions for red and blue violations, the excluded counties, the update-timing and county-variation limits, and the grading-system statement.
• Oswego County Health Department, Food Service Inspections — what critical (red) and non-critical (blue) violations cover, and the expectation that critical violations are corrected during the inspection. County health departments conduct these inspections under the state programme; the state’s own regulation page, linked from the dataset, no longer resolves.
• Chain figures are computed from the same committed snapshot as the New York State page; the method is on the methodology page.

New York State data: The analyses, interpretations and conclusions drawn from New York State data on this site were reached by FastFoodFact and not by the New York State Department of Health. The department states that this data was not collected for use as a restaurant grading system and does not support or endorse such a use; the figures here describe groups of inspections and are not a grade of any restaurant.

Informational only. This is a reading of public inspection records, not food-safety, legal or dining advice, and no figure here is a grade of any individual restaurant. A figure describes what inspectors recorded across a set of visits in one place over one period — not a brand, and not the location near you.

Last reviewed: September 2026 · Against the New York State Department of Health's published inspection dataset and its stated limitations.