Critical Control Points Explained: How to Identify and Monitor CCPs

Every food business operating in Ireland or the UK is required to have a permanent procedure based on HACCP principles. In practice, most of the argument at inspection comes down to one question: which steps in your process are critical control points, and can you prove you controlled them?

This guide covers what a CCP actually is, how to tell one from a prerequisite programme, the critical limits you are likely to be held to, and what monitoring has to look like to survive an audit. It sits underneath our overview of the seven principles of HACCP, and goes deeper on principle two.

What is a critical control point?

A critical control point is a step where control is essential to prevent, eliminate or reduce a food safety hazard to an acceptable level.

The word doing the work is essential. If the hazard can be controlled at a later step, the step you are looking at is not a CCP. Cooking is the classic example. For most cooked products nothing downstream will destroy surviving pathogens, so cooking is the last line of defence and therefore critical.

A CCP has four things attached to it:

• A hazard it controls
• A critical limit, meaning a measurable value separating safe from unsafe
• A monitoring procedure covering who checks, how often, and by what method
• A corrective action for when the limit is breached

If any of those four is missing, you do not have a controlled CCP. You have an intention.

CCP or prerequisite? The distinction auditors test

This is where most HACCP plans fall down, in both directions.

Prerequisite programmes are the conditions that make safe food production possible at all: cleaning schedules, pest control, staff hygiene, maintenance, supplier approval, waste management. They apply across the whole operation rather than at one identified step.

Critical control points are specific steps where you apply a measurable control to a specific hazard.

Two failure modes:

• Too many CCPs. Every cleaning task becomes a critical control point, the plan becomes unmanageable, and monitoring degrades because nobody can keep up with it.
• Too few CCPs. A genuine control step gets treated as routine good practice, so nothing is measured and nothing is recorded.

How to identify your CCPs

Work through your process flow diagram step by step. At each step, ask in order:

1. Is there a hazard at this step that needs controlling?
2. Is there a control measure in place for it?
3. Is this step specifically designed to eliminate or reduce that hazard to an acceptable level?
4. Could contamination occur at, or increase beyond, acceptable levels here?
5. Will a later step eliminate or reduce the hazard to an acceptable level?

If a later step handles it, this one is not a CCP. If nothing downstream handles it, it is.

Record the reasoning, not just the conclusion. An auditor asking why chilling is a CCP at one site and not another wants the determination, and "it always has been" is not an answer.

Critical limits: the numbers you will be held to

Critical limits must be measurable, and must be validated against your process rather than adopted from a template. That said, these are the values most Irish and UK food businesses work to:

Equivalent time and temperature combinations are acceptable for cooking where you can demonstrate the same lethality, such as 70°C for two minutes. If you rely on an equivalent, keep the validation with the plan.

Note what is not in that table: anything in Fahrenheit, and anything from the FDA Food Code. If your plan cites US figures, an Environmental Health Officer will ask why.

Common CCPs by sector

The hazards differ by what you make. Some patterns:

Bakery

• Baking core temperature, where the product carries a microbiological hazard
• Allergen segregation, usually managed as a prerequisite through production sequencing and clean down, but treated as a CCP where cross contact cannot otherwise be controlled
• Retarding and proving, the reading nobody takes because it happens overnight

See how this works in practice on our bakery page.

Deli and food service

• Hot holding across the full service period, not checked once at the start
• Chilled display, since cabinets opened constantly behave nothing like a sealed fridge
• Cooling of cooked product through the danger zone

More on deli controls.

Meat and butchery

• Cooking, where you produce ready to eat product
• Chilling and cutting room temperature, a process that generates heat in a room that must stay cold
• Separation of raw and ready to eat, by space, by time and by dedicated equipment, each recorded

More on butchery controls.

Food retail

• Chilled and frozen storage across a large estate of units
• Date code control and removal of out of date stock
• Goods in temperature at delivery

More on retail controls.

Ready to eat production

RTE products carry the strictest controls, because there is no kill step between your process and the consumer:

• Pasteurised or otherwise treated ingredients only
• Environmental monitoring for pathogens, particularly Listeria
• Physical separation of raw and RTE materials and equipment
• Metal detection or other foreign body control
• Validation that the lethality step actually delivers the reduction claimed

Monitoring: the part that fails in practice

A critical limit with no monitoring is decoration. Monitoring has to specify:

• What is measured
• How, whether by probe, sensor, visual check or calibrated instrument
• How often, whether every batch, hourly or continuously
• Who, meaning a named trained person rather than "the kitchen"

Some practices that hold up:

• Use calibrated instruments, and record the calibration
• Record the actual reading rather than a tick. A figure of 12°C tells an auditor something. A tick tells them nothing
• Keep the record at the point of the check, not written up later from memory
• Rotate responsibility across several trained people, so a holiday does not create a gap
• Review the records for trends, not only for breaches

Corrective actions

Define these in advance, at each CCP, before anything goes wrong. A corrective action needs to cover two things:

1. The product. What happens to the food affected, whether that is continue cooking and re probe, move to another unit, discard, or hold pending assessment.
2. The process. What stops it happening again, such as an engineer called, a procedure changed, or staff retrained.

Both get recorded. In paper systems the product decision usually gets written down and the process action does not, which is exactly the half an auditor asks about.

Verification and records

Monitoring proves the limit was met on the day. Verification proves the system works: QA review of completed records, internal audits, calibration checks, and scheduled review of the plan itself, annually or whenever a process, product or supplier changes.

Records are the seventh principle and the first thing requested at inspection. They need to be accurate, dated, attributable to a named person, and available on request. There is no requirement for them to be on paper, and digital check sheets meet those tests more reliably than a clipboard does.

The moment those records matter most is a withdrawal. See food recalls and withdrawals for what Article 19 requires once something has gone wrong.

Where software fits

None of the above requires software. A small operation can run a compliant HACCP system on paper, and plenty do.

What software changes is retrieval and consistency. The plan stays linked to the records that evidence it, so changing a critical limit changes the check sheets monitoring it. Corrective actions are prompted at the moment of a breach rather than written up later. Records filter by site, sheet and date range instead of being hunted through folders.

That is the difference our customers describe. Greenaway Foods found audit preparation became a search function rather than a fortnight of digging. Read the case studies, or see how HACCP software covers all seven principles.

Frequently asked questions