Monitoring is a planned sequence of observations or measurements at a critical control point to assess whether it is under control and to produce an accurate record for later verification. HACCP Principle 4 requires every CCP to have a monitoring procedure that specifies what is measured, how, when, and by whom, using calibrated instruments and records made as the event happens.

Monitoring is where the plan meets the shift. A perfect critical limit is worthless if nobody checks it, checks it late, or checks it with a thermometer that reads five degrees high. This post covers the four questions every monitoring procedure must answer, the difference between continuous and periodic monitoring, why instruments have to be calibrated, and why records made at the time of the event are the whole point. It follows directly from Principle 3, critical limits because monitoring is how you find out whether a limit is being met.

What is monitoring in HACCP?

Monitoring is the act of measuring or observing a CCP against its critical limit on a planned schedule, and recording the result. NACMCF defines it as "a planned sequence of observations or measurements to assess whether a CCP is under control and to produce an accurate record for future use in verification." It does three jobs at once: it tracks the process so operators can spot drift, it signals the moment a critical limit is breached so corrective action can start, and it creates the written record that verification and auditors later review.

That third job is easy to underrate. Monitoring is not just watching; it is watching in a way that leaves proof. A cook checked and passed but never recorded did not, as far as an auditor is concerned, happen. This is why monitoring, record-keeping, and the eventual HACCP certification audit are so tightly linked, and why the design of the monitoring record matters as much as the measurement itself.

What are the four questions every monitoring procedure answers?

Every monitoring procedure must answer four questions: what is monitored, how it is monitored, when or how often, and who is responsible. Leave any one vague and the procedure fails in practice, because a gap in any of the four is where deviations slip through unrecorded.

The four questions a monitoring procedure must answer A monitoring procedure answers four questions WHAT the parameter & its critical limit e.g. internal temp ≥ 165°F, detector rejects ≥ set test piece HOW the method & instrument e.g. calibrated probe in the thickest part; inline data logger WHEN the frequency e.g. every batch, every hour, or continuously with alarm WHO the trained, named person e.g. line operator, QA tech, who signs or initials the record
What, how, when, who. The fourth box is in red because named accountability is the one plants most often leave implied, and it is the first thing an auditor pins down.

What is the parameter and its critical limit: internal temperature at or above the cook limit, a metal detector rejecting on a defined test piece, a pH reading at or below 4.6. How is the method and the specific instrument, placed where the measurement is meaningful, for example a calibrated probe in the thickest part of the product. When is the frequency, chosen so a deviation is caught before much product is affected. Who is a trained, named person who performs the check and signs or initials the record. USDA's rule reinforces the fourth point: monitoring entries must be signed or initialed by the employee who makes them.

Worked concretely, a cook-step procedure reads like this: what is the internal temperature of the coldest-looking unit against a limit of 165°F; how is a calibrated probe thermometer inserted into the thickest part; when is the first unit of every batch plus one every 30 minutes; who is the cook-line operator, backed up by the QA technician on break. Written that tightly, there is nothing left to interpret on the floor, and the record that comes off it is unambiguous. Vague procedures ("check temp regularly") are the ones that generate blanks, disputes, and audit findings.

What is the difference between continuous and periodic monitoring?

Continuous monitoring records the parameter at every moment, usually with an instrument, while periodic monitoring checks at set intervals. The choice depends on the CCP: some parameters can be watched constantly by a device, others only sampled by a person, and the frequency of periodic checks has to be tight enough to limit how much product is exposed if a deviation is found.

Continuous versus periodic monitoring Continuous vs periodic monitoring Continuous every moment recorded · alarm on breach Periodic ✓ ✓ ✗ ✓ ✓ deviation in a gap → hold all product back to the last good check
Continuous monitoring pinpoints a breach to the moment. Periodic monitoring only tells you it happened sometime since the last good check, so everything in that window is suspect. Frequency is a product-at-risk decision.

Continuous monitoring, a thermal recorder on a cooker or an inline metal detector, is the gold standard because it captures every moment and can alarm the instant a limit is breached. But even continuous instruments need periodic human checks to confirm they are working and to review the record. Periodic monitoring is fine for many CCPs, but the interval determines your exposure: if you check pH every two hours and find a deviation, every unit made in that two-hour window is suspect and must be held. That is the direct link to Principle 5, corrective actions and it is why tighter monitoring frequencies reduce the amount of product ever put at risk.

Why do monitoring instruments have to be calibrated?

Monitoring instruments have to be calibrated because a reading is only as trustworthy as the device that produced it, and an out-of-calibration instrument silently invalidates every check it took. A thermometer reading 5°F high tells you a cook passed when it may have failed, so calibration is what makes monitoring data mean anything. USDA's HACCP rule lists calibration of process-monitoring instruments as a required ongoing verification activity.

Calibration is itself a documented program: known frequencies, reference standards, records of each calibration, and a defined response when an instrument is found out of tolerance. That last part matters, because if an instrument is found reading wrong, you have to reach back and question every measurement it made since its last good calibration, which can put product on hold. A well-run calibration program, covered in calibration for food safety is a prerequisite for monitoring to be believable at all.

The same trust question applies to the environment around the CCP, not just the instrument at it. A cook that hits temperature but a post-cook zone that harbors Listeria still ships an unsafe product, which is why monitoring at CCPs sits alongside an environmental monitoring program rather than replacing it. Principle 4 proves the control step worked; the surrounding prerequisite monitoring proves nothing recontaminated the product afterward. Auditors read the two together.

How do you build a monitoring procedure for a CCP?

You build a monitoring procedure by working through the four questions in order, then pinning down the record and the frequency against how much product a deviation would put at risk. The sequence below produces a procedure an operator can follow and an auditor can verify.

  1. Restate the CCP and its critical limit. Name the step, the hazard, and the exact critical limit being monitored, so the target is unambiguous.
  2. Define what is measured. Specify the parameter and the pass/fail value, including where on the product or line it is measured.
  3. Define how, and on what instrument. Name the method and the specific calibrated instrument, and where and how it is applied.
  4. Set the frequency, sized to the risk. Choose continuous or an interval tight enough that a deviation exposes an acceptable, recoverable amount of product.
  5. Assign a trained, named person. State who monitors and who is the backup, and confirm they are trained on the procedure and the instrument.
  6. Design the record. Build a record that captures the actual value, the date and time, and the monitor's signature or initials, made at the time the check occurs, per USDA record rules.

Facts worth pinning

Where monitoring actually breaks

Monitoring rarely fails because the plan is wrong; it fails because the record is thin. Pre-op sheets filled at the end of the shift from memory, temperature logs with suspiciously identical values, checks that happened but were never written down, an instrument nobody calibrated for six months. Every one of those is a real finding, and every one is a records problem more than a food-science problem. The fix is to make the record trustworthy by construction: timestamped at the moment of the check, tied to a specific calibrated instrument, with required fields that cannot be left blank.

That is exactly the digitize-the-paper move Harmony runs for production and quality logs (see how CLS did it): CCP checks on tablets, captured as they happen, so the monitoring record is honest and audit prep becomes a query instead of a binder hunt. It connects monitoring straight into Principle 6, verification where those same records get reviewed, and into your HACCP plan as living evidence. See how a connected operation keeps monitoring data live on the features overview.