What Restaurant Inspection Grades and Scores Actually Mean

Ask someone what a restaurant's health grade means and you will usually get a confident answer. Ask them what the grade is out of, who set the cut-off, and whether it would be the same grade one state over, and the confidence drops fast.

That is not the public being careless. It is a genuinely fragmented system. The United States has no single inspection standard, no single scoring method and no single way of publishing the result.

Why there is no national standard

The Food and Drug Administration publishes the Food Code, a model set of rules for retail food safety, currently in its 2022 edition (FDA Food Code). It is genuinely good and very widely used. It is also, importantly, a model and not a mandate. States choose whether to adopt it, which edition to adopt, and what to change on the way in.

So one state may be working from the 2022 edition, its neighbour from an earlier one, and a third from its own state code that borrows heavily from the model without matching it. The underlying food safety science is broadly shared. The rule numbers, the scoring weights and the thresholds are not.

The four formats you will encounter

Letter grades

Only a handful of jurisdictions use a posted A/B/C style grade. They are the most publicly legible format by a wide margin, which is exactly why people over-read them. A grade is a compression of a scoring system that was designed locally. Two cities can both hand out an A and mean different things by it.

What a letter grade does tell you reliably: whether the operator cleared the local bar on the day of the visit. What it does not tell you: how close they were to the line, or what the specific findings were.

Numeric scores

More common than grades. Usually a starting figure with points deducted for each violation, weighted by how dangerous the violation is. The weighting is where jurisdictions diverge most, so comparing a score across state lines is close to meaningless.

Within one jurisdiction, though, scores are useful — particularly the trend. Three consecutive visits moving up is a real signal. One good score in isolation is not.

Pass / fail, or "in compliance"

A binary outcome. On its own it is thin, but these reports almost always carry a full narrative underneath, and the narrative is where the actual information lives.

Narrative reports only

The most common format nationally. No grade, no score — just the inspector's observations, each tied to the rule it relates to, usually with a note on whether it was corrected on site.

People find these frustrating because there is no number to latch onto. They are in fact the most honest format available. You are reading what the inspector actually saw.

The distinction that matters more than the score

Nearly every system separates violations into two tiers, whatever it calls them.

Risk factor violations — sometimes called critical, priority, or foodborne illness risk factors — are the practices directly linked to people getting sick. Food held at the wrong temperature. Inadequate cooking. Cross-contamination. A sick employee working a shift. Poor hand hygiene. The FDA has studied the prevalence of exactly these behaviours in retail food establishments over successive data collection periods (FDA Retail Food Risk Factor Study).

Good retail practices are everything else — a cracked floor tile, an unlabelled container, a missing thermometer in a dry store. They matter, they should be fixed, and they very rarely make anyone ill on their own.

Here is the practical consequence. A restaurant can accumulate a mediocre-looking score from a pile of good-retail-practice findings while having a genuinely safe kitchen. Another can post a respectable score with a single temperature-abuse finding buried in the narrative — and that one finding is the one that puts people in hospital.

The number tells you how the day went. The violation list tells you whether the kitchen is safe. If you read only one part of a report, read the violations.

Reading your own record as an operator

Two habits are worth building.

First, sort your findings into the two tiers yourself. If everything on your report is a good retail practice, you have a housekeeping problem and a clear fix list. If there is a risk factor finding, that is a process problem, and cleaning harder will not solve it.

Second, watch for repeats. A violation that appears on two consecutive reports reads very differently to an inspector than the same violation appearing once. Repeats are what turn a correctable report into an enforcement file. Closing out a repeat is the highest-value thing you can do between visits.

Run the ten-minute self-check — it walks the same ground an inspector covers, in the order they usually cover it.