Critical vs. Non-Critical Violations: How to Read the Two Halves of an Inspection Report

Pull up any restaurant's health inspection report and you'll notice something quickly: the violations are not treated as equals. One line says the establishment was cited for a hand sink without soap. Another says the walk-in cooler was holding chicken at 51°F. A third says a light shield was cracked in the dry storage room.

To a customer skimming the page, that's just three violations — three strikes. To a health inspector, those three lines live in completely different tiers, and only one of them describes something that could have made somebody sick that day.

Understanding that split is the single most useful skill for reading an inspection report. It's the difference between "this place has eleven violations, avoid it" and "this place has eleven violations, ten of which are paperwork and lighting, and it's probably fine." Here's how the tiers actually work.

The FDA Food Code split: priority, priority foundation, core

Most U.S. health departments build their inspection forms on the FDA Food Code, a model regulation the FDA updates every few years. States and counties adopt it, amend it, and enforce it locally — which is why the labels differ from place to place — but the underlying three-tier logic is remarkably consistent.

Priority items are the ones tied directly to preventing foodborne illness. Cooking temperatures. Cold-holding and hot-holding temperatures. Cross-contamination between raw and ready-to-eat food. A sick employee working a food-handling station. Approved sources for shellfish and other high-risk items. Sanitizer concentration in the warewashing machine. If a priority item is out of compliance, there is a plausible, direct path from that failure to somebody getting sick.

Priority foundation items are one step back. They're the supports that make priority compliance possible. A hand sink with no soap or paper towels isn't itself contamination — but it makes proper handwashing impossible, and handwashing is a priority control. No working thermometer in the cooler. No date-marking system on prepped food. No certified food protection manager on staff. These are the conditions that let priority failures happen.

Core items are everything else: maintenance, cleanliness of non-food-contact surfaces, physical facility upkeep, general sanitation, signage. A cracked floor tile. A dusty vent hood grate. A missing light shield. Peeling paint in the back hallway. These matter — they're in the code for a reason, and a restaurant that lets enough of them pile up is usually letting other things slide too — but no core violation on its own is going to make anyone sick tonight.

The local vocabulary problem

Here's where it gets confusing for anyone comparing restaurants across cities. The tiers are broadly the same; the words are not.

• "Critical" and "non-critical" — the older FDA terminology, still used by a large number of jurisdictions and still the phrasing most people search for. Roughly: critical maps to priority plus priority foundation, non-critical maps to core.
• "Priority / priority foundation / core" — the current FDA Food Code language, used by jurisdictions that have adopted newer code editions.
• "Red / blue" — some departments color-code the form itself, with red items being the health-risk tier.
• "Public health interventions" and "good retail practices" — how the inspection form is often physically divided into two sections.
• "Foodborne illness risk factors" — the header on the section covering the five CDC-identified risk factors: food from unsafe sources, inadequate cooking, improper holding temperatures, contaminated equipment, and poor personal hygiene.

If you're reading a report and can't tell which tier a violation falls into, look at the section headers rather than the violation text. Nearly every inspection form separates the two halves visually, and the risk-factor half is almost always printed first.

Why point values don't map cleanly onto tiers

Numeric-score jurisdictions add another wrinkle. In a 100-point system, a violation's point deduction is supposed to reflect its risk weight — critical items deduct more, core items deduct less. But the weighting schemes are local, and they don't line up between cities.

A temperature-abuse violation might carry a different deduction in one county than in the next one over. Some systems increase the deduction when the same critical item was cited on the previous inspection. Some cap the total deduction for repeated core items. Some let an operator correct a violation on the spot during the inspection and reduce or waive the deduction entirely — which means two restaurants with similar underlying conditions can post very different scores depending on whether the manager fixed it while the inspector was standing there.

This is why comparing raw scores across city lines is mostly meaningless, and why the tier breakdown is more informative than the number at the top of the page. A 92 with two priority violations is a worse report than an 88 with six core violations, even though the number says otherwise.

"Corrected on site" is doing a lot of work

One of the most commonly misread fields on an inspection report is the correction status. Most reports mark each violation with something like corrected on site, corrected during inspection, or a compliance-by date for follow-up.

"Corrected on site" means the inspector observed the problem, the operator fixed it before the inspector left, and the inspector verified the fix. For a lot of violations, that's genuinely the end of it — the soap dispenser got refilled, the sanitizer bucket got remixed to the right concentration, the raw chicken got moved below the produce.

But the phrase describes the fix, not the cause. A cooler holding food at 51°F can be "corrected on site" by moving the food to a working cooler and discarding what's been out too long. That's the right immediate action. It does not mean the cooler is repaired, and it does not tell you whether the same cooler was cited three months ago. The correction status is a snapshot; the pattern across inspections is the actual signal.

Reading the pattern instead of the snapshot

Any single inspection is a short window on one specific day. Inspectors work with what they can observe in that window, and a lot of what determines a restaurant's real food-safety posture — staff training, manager turnover, whether the walk-in has been failing intermittently for a year — isn't visible in a single visit.

The pattern across several inspections is far more informative. When you're looking at a restaurant's history, these are the questions worth asking:

• Are priority violations repeating? The same temperature violation across three consecutive inspections is a system problem, not a bad day. A different core violation each time is normal operational noise.
• Does the count spike and then settle? A bad inspection followed by a clean re-inspection usually means the operator took it seriously. A bad inspection followed by another bad one usually means they didn't.
• What triggered the visit? Many departments note whether an inspection was routine, a follow-up, or complaint-driven. A complaint-driven inspection with findings reads very differently from a routine one.
• How recent is the most recent one? A glowing report from two years ago tells you about a kitchen that may have changed managers twice since.
• Did anything escalate? Suspension, closure, embargo of product, or a hearing referral are all several steps beyond a normal citation.

What a clean report doesn't prove

It's worth being honest about the ceiling here. A zero-violation inspection is a good sign and not a guarantee. Inspectors check a defined list of observable conditions during one visit. They aren't auditing supplier chains, they aren't watching the line during a Saturday rush, and they can't see what happens after they leave.

Inspection data is one input. It's a genuinely useful one — it's standardized, it's produced by trained public employees with no commercial stake in the restaurant, and it's public record — which is more than you can say for most restaurant information online. But it works best as a filter for obvious problems rather than as a seal of approval.

The practical version

If you only remember one thing: read the tiers, not the total.

Open the report. Find the section break between risk-factor violations and general retail practices. Count each side separately. Then check whether anything on the risk-factor side shows up again on the previous inspection. That's a ninety-second read, and it tells you more than the letter grade or the score ever will.

Two violations on the risk side deserve more attention than nine on the maintenance side. A repeat on the risk side deserves more attention than either. And a restaurant with a long, boring history of small core violations and nothing else is, statistically, roughly what a normal well-run kitchen looks like on paper.

About the author — Rick Jenkins is the founder of AI Revenue Forge. ARF builds vertical-specific AI virtual receptionists for service businesses in HVAC, dental, medspa, real estate, home health, credit repair, and pawn shops. Headquartered in Charlotte, NC. Part of Jenkins Worldwide Enterprises.