Notes / Reading a record

How to Read a Restaurant Inspection Report Line by Line

Once you have the record, the useful work is reading it. Here is how to parse grades, scores, and narrative findings without mistaking a housekeeping list for a safety verdict.

Finding the record is half the job. Looking it up gets you to the right portal. Reading what you found is the other half — and it is where most people stop too early, at the letter or the number on top.

A restaurant inspection report is a photograph of one visit. It is not a running average, and it is not a national grade. What follows is a line-by-line habit for reading any format you are likely to meet in the United States.

Start with the format, not the verdict

Jurisdictions publish results in different shapes. They are not interchangeable.

  • Letter grade. An A/B/C (or similar) card. Easy to read, easy to over-read. The cut-offs behind the letter are local.
  • Numeric score. Usually points deducted from a starting figure. A 92 here and a 92 one state over can describe different kitchens.
  • Pass / fail or “in compliance.” Binary on the surface. The detail lives in the narrative underneath.
  • Narrative only. No grade or score — just observations coded to rules. The most common format nationally, and the most informative if you read it.

If the portal shows a grade or score and a violation list, treat the list as the primary document. The summary is a compression. Compression loses information on purpose.

Separate illness-risk findings from housekeeping

Almost every modern code splits findings into two tiers, whatever the local labels.

Risk-factor findings — sometimes called critical, priority, or major — track practices tied to foodborne illness: handwashing, bare-hand contact with ready-to-eat food, cross contamination, sanitizing food-contact surfaces, cold and hot holding, cooling, date marking, cooking, reheating. Those are the ten practices the FDA measures nationally.

Good retail practices cover floors, walls, lighting, general sanitation, and similar maintenance. They matter. They should be fixed. They very rarely make anyone ill on their own.

Sort every line on the report into one of those two piles before you decide what the visit means. A long report full of housekeeping can describe a safer kitchen than a short report with one live temperature finding.

Read each violation line for four things

Most narratives, whatever the portal design, give you the same four pieces of information if you look for them.

  1. What was observed. The facts of the moment — a cooler reading, a hand sink blocked, raw chicken above ready-to-eat product.
  2. Which rule it maps to. A code citation or category. Useful when you compare visits, because the same rule repeating is the signal.
  3. Whether it was corrected on site. Many reports note COS or “corrected during inspection.” That matters for the day’s outcome. It does not erase the finding from the pattern across visits.
  4. Whether it is a repeat. Some portals mark repeats explicitly. If they do not, open the last two or three reports yourself and match the rule numbers.

Corrected on site is good operational news. It is not the same as “never happened.” Operators who only celebrate COS and never fix the process see the same line again next time.

One visit is a photo; three visits are a pattern

A single report describes one kitchen on one day — often a busy service day, sometimes a quiet one. Kitchens change between visits for better and for worse.

Three consecutive reports tell you more than any score. Ask:

  • Do the same risk-factor rules reappear?
  • Are COS notes covering the same process failure repeatedly?
  • Is the housekeeping list shrinking while risk factors stay open — or the reverse?

Repeats are what change how an inspector reads a file. A first cold-holding citation and a third cold-holding citation are not the same event, even when the temperature reading looks similar.

What a grade or score cannot tell you

It cannot tell you how close the kitchen was to the cut-off. It cannot tell you whether the weight came from illness-risk findings or from paint and tile. It cannot be compared across a county or state line, because the deduction tables and posting rules differ.

It also cannot tell you what happened after the visit. A temporary closure followed by clean re-inspections is a different story from a closure that repeats. Read the follow-up reports before you treat a closure as a permanent verdict.

If you run the restaurant

Print or save the last three reports. Highlight every risk-factor line. Assign a named owner to each open item. Photograph corrections with dates. Then walk the kitchen with the ten-minute self-check on a normal service day — the same conditions under which the next unannounced visit will land.

If the latest report is bad, use the first-48-hours sequence rather than arguing the score in the abstract.

If you are deciding where to eat

Prefer the violation list over the letter. Prefer recent pattern over a single old mirror site. Prefer the issuing health department over a third-party app that may not have refreshed. And remember that posting rules differ — in some places the result is already on the door.

Look up the official portal for your state, then read the lines — not only the summary on top.