Restaurant Corrective Action Guide: What to Do When a Food Safety Check Fails

A corrective action is what you do when a food safety check fails: the walk-in reads 45°F, the sanitizer bucket tests at zero, the chicken salad has no date, the cook skipped the handwashing sink. It has two halves. The immediate half makes the food and the guest safe right now (discard, reheat, re-chill, re-wash). The preventive half finds out why it happened and changes something so it does not happen again. Most kitchens do the first half reasonably well and almost never do the second. This guide explains both, shows what a usable corrective-action record looks like, and lays out a simple process a manager can run every shift.

What counts as a corrective action

In HACCP terms, corrective action is the fifth of the seven principles: the pre-planned response when monitoring shows that a critical limit was not met. FDA’s HACCP Principles and Application Guidelines (opens in a new tab) describe corrective actions as procedures that determine and correct the cause of the deviation, determine the disposition of the affected product, and record what was done. The FDA Food Code brings the same idea into everyday retail operation through “active managerial control”: the person in charge is expected to monitor the risk factors, recognize when something is out of control, and fix it (see Food Code Annex 4).

You do not need a formal HACCP plan to need corrective actions. Every temperature log, line check, opening checklist, or self-audit produces failures, and each failure needs a documented response. The difference between a kitchen that passes inspections and one that does not is rarely whether problems occur; it is whether the problems get a response that an inspector can see.

Common triggers in a restaurant

Temperatures and times in the table follow the current FDA Food Code (opens in a new tab) (§3-501.16 holding, §3-501.14 cooling, §3-501.17 date marking) and the USDA FSIS safe temperature chart (opens in a new tab). Your jurisdiction may have adopted an older edition or different numbers; check before you print a procedure.

Immediate action and preventive action are different jobs

Immediate action: protect the food and the guest

The immediate action is decided in advance and executed by whoever finds the problem. It should never require a phone call to figure out. Write the response next to each check on your logs and checklists so a line cook at 7 a.m. knows exactly what to do with a 46°F reach-in without waiting for a manager. The two most common immediate actions are “discard” and “re-process” (reheat, re-chill, re-wash, re-sanitize), and the decision rule is almost always time: how long has the product been out of the safe range, and do you actually know?

Preventive action: change something so it stops recurring

Preventive action is the manager’s job and it happens after service. It asks why the failure occurred and changes a process, a piece of equipment, a schedule, or a training gap. A reach-in that runs warm every Friday lunch does not need a fourth service call; it needs a smaller load or a second unit. A station that misses date labels every night needs labels at the station and a closing step that checks for them, not another reminder in the pre-shift meeting.

The simplest root-cause tool that works in a kitchen is asking “why” a few times until you reach something you can change:

Finding:  Line reach-in #2 at 46°F at 1:40 p.m. (second time this week)
Why?      Compressor was running constantly and could not keep up.
Why?      Unit was packed to the top with backup product for the lunch rush.
Why?      Prep list calls for a full day of proteins to be staged at 10 a.m.
Why?      Walk-in is 40 feet away and the line doesn't want to run back mid-rush.
Why?      No mid-shift restock step exists; staging everything at once is the workaround.

Change:   Add a 12:30 restock task to the shift routine; cap the reach-in at
          two-thirds full; re-check unit at 1:00 and 2:00 for two weeks to verify.

What a corrective-action record should contain

A corrective-action entry is only useful if someone else can understand it later: a regional manager, a new GM, a health inspector, or an attorney. Every entry should answer the same questions:

• What failed, where, and when — the check, the unit or station, the reading, the time, and who observed it.
• Product affected — item names and quantities, and what happened to them (discarded with weight or count, reheated to a recorded temperature, re-labeled).
• Immediate action taken — in plain language, with the time.
• Root cause — the best explanation you have, even if provisional.
• Preventive action — the specific change, who owns it, and the due date.
• Verification — how and when someone will confirm the fix worked, and the result.
• Evidence — photos of the reading, the product, the repair invoice; a manager’s initials or signature.

CA #2026-0914-03                    Store 07   Opened: 09/14  07:12   By: D. Okafor (AM)
TRIGGER    Opening temp check — walk-in cooler air 44°F; product: sliced turkey 44°F,
           cooked rice 43°F, diced tomato 42°F. Last good reading 09/13 22:05 (38°F).
IMMEDIATE  07:15 All TCS product with unknown time > 41°F discarded:
           turkey 6.2 lb, rice 2 hotel pans, chicken salad 1 pan, dairy (see waste log).
           Produce/whole items retained. Unit door closed, thermostat verified at 36°F set.
           07:30 Service call placed (ticket 55812). Product staged in reach-ins #1, #3.
ROOT CAUSE Evaporator fan iced over; door found propped open at close per camera (22:40–23:10).
PREVENTIVE 1. Closing checklist: add "walk-in door closed & latched — photo" (owner: GM, due 09/15)
           2. Defrost cycle inspected by tech; fan replaced (owner: GM, done 09/14 11:20)
           3. Retrain closing crew on door discipline at 09/16 pre-shift (owner: D. Okafor)
VERIFY     Walk-in checked hourly 09/14–09/15: all readings 35–38°F. Photo IMG_2210–2216.
           Closing photo present 09/15–09/21 nightly. Closed 09/22 by GM (initials: R.P.)

Notice that the entry links the failure to two existing systems: the closing checklist (which gets a new step) and the temperature log (which supplied the last-good reading). That is what makes it a control instead of a story.

A corrective-action process a manager can actually run

1. Pre-write the immediate actions. Every check on your logs and checklists gets a one-line “if it fails” instruction. The HACCP checklist in this resource center shows how to attach these to critical limits.
2. Make it easy to raise one. Staff should be able to open a corrective action from the failed check itself, on paper or on a phone, in under a minute. If it takes a form from the office, it will not happen.
3. Assign an owner and a due date, every time. An unowned corrective action is a note. Preventive actions that require money or a vendor go to the GM or above; everything else stays with the shift lead.
4. Review open items daily. Two minutes at the start of the manager’s day: what is open, what is overdue, what closed yesterday. The list should be short. If it is long, you have a system problem, not a discipline problem.
5. Verify before closing. Closure means someone confirmed the fix works, not that the task was marked done. For equipment, that usually means a period of readings in range; for behavior, a period of observed compliance.
6. Look for patterns monthly. Group closed items by trigger, station, shift, and person. Three door-left-open findings on Tuesday closes is a scheduling or training issue with a name on it.

Who can approve disposal and re-processing

Decide in advance who may authorize discarding product and who may authorize reheating or re-serving. Many operators let any employee discard (erring toward safety) but require a manager to approve re-processing, because that is the decision with risk attached. Whatever you choose, write it down and make it part of onboarding.

Records, retention, and the inspection

The FDA Food Code requires records for specific situations (for example, when a variance or HACCP plan is in place under Chapter 8, or for time-as-a-public-health-control procedures under §3-501.19) and gives the regulatory authority the right to review them. For routine logs and corrective actions in an ordinary restaurant, the Food Code does not prescribe a universal retention period; retention rules, where they exist, come from your state or local code, your franchisor, or your insurer. A practical default many operators use is to keep at least the current and previous inspection cycle’s records on hand and archive older ones. Confirm the requirement with your health department rather than assuming.

Common mistakes

• Treating the immediate action as the whole job. Discarding the product protects today’s guests. It does nothing for tomorrow’s.
• Root causes that blame a person. “Cook forgot” is a symptom. Why was forgetting possible? What would make it hard to forget?
• Corrective actions that live in text messages. They cannot be searched, assigned, tracked, or shown to an inspector, and they vanish with the phone.
• No verification step. A gasket replacement that did not fix the unit is a closed task and an open problem.
• No pattern review. The value of the record compounds only when someone reads it in aggregate.
• Skipping the waste record. Discarded product is food cost. Quantities on the corrective action feed the waste log and make the business case for the preventive fix.

Paper, spreadsheets, or software

Paper works for a single location with a disciplined manager. It breaks when you need to assign work to someone on another shift, prove a photo was taken at a certain time, or see open items across several stores. If you are evaluating software for this, our restaurant checklist software buyer’s guide and food safety software guide cover what to look for. The specific capabilities that matter for corrective actions are: raising a task directly from a failed check, assigning it to a named person with a due date, requiring photo evidence to close it, alerting a manager when a critical check fails, and reporting open and overdue items across locations. For a broader view of how these tools fit together, see the restaurant operations software hub.