Home » Blog » Codex 2023 CCP Decision Tree

Codex 2023 CCP Decision Tree

Reissued Codex 2023

You may recall that in the 2020 update of the Codex HACCP Guideline, the CCP decision tree was removed for updating, it has now been amended and reissued.

Firstly, don’t panic! Especially if you’re just about to do a HACCP course and exam – it will take a while before the awarding bodies integrate this into their exam questions!

If you’re managing a HACCP system, then there will be some actions, but again this new version of the tree will take some time to filter into common use. You may already have chosen to use the widely accepted Camden BRI version of the CCP decision tree, in which case, you could choose to stick with that.

Here’s a link to the Codex Guidelines that includes the new version of the tree, the decision tree is on page 47.

I’ve given my thoughts on the new version below as both headlines and detail. I’d be interested to hear any comments you have as well. Please post your comments below.

The headlines:

  • This new CCP decision tree from Codex combines and reorders the previous Codex version and the more practically used 5 question CCP decision tree from Campden BRI (Campden BRI, HACCP: a practical guide, 5th edition, Guideline 42)
  • The new 2023 version still has 4 questions. The order of some questions has changed. The sub-question of 1A ‘is control necessary at this step for safety’ has been removed
  • Q1 enables significant hazards that are controlled by PRPs to exit the tree at this early stage, removing the risk of creating unnecessary CCPs, as we had with the previous version (this was the additional element that the Campden version brought us)
  • Both Q2 and Q4 refer to the specificity of the control measure/step to control the significant hazard – this is a key difference from the previous version
  • Q3 is very similar to the previous Q4, focusing on whether a subsequent step will control the hazard. The wording has been expanded here to include ‘prevent’, which is more comprehensive than the previous version
  • Q4 is what was Q2 in the previous version, again with the addition of the word ‘prevent’
  • As always, we are only putting significant hazards through the CCP decision tree

In my opinion, the main benefit is the new Q1, which reflects the Campden BRI version. It’s unlikely that the new tree will change any of your CCP decisions as these are by now well-established.

Codex 2023 decision tree

The detail:

Question 1

‘Can the significant hazard be controlled to an acceptable level at this step by prerequisite programmes (e.g. GHPs)?’

The start of the new tree is more in line with the Campden version and provides an opportunity to say, at the outset, that a significant hazard is managed by a PRP.

Answering ‘yes’ to Q1…
If you answer ‘yes’, then the tree tells you it’s not a CCP and the hazard won’t progress any further through the tree as it is managed by PRPs (we should have stated the specific control measure, and the PRP its sits within, in our hazards analysis table). Do keep in mind that this can be quite an arbitrary question and we must ensure that the relevant PRP does actually control the significant hazard (has the significant hazard been considered in the PRP and does the PRP include a control measure for the identified significant hazard?) In most larger businesses many significant hazards will be controlled by PRPs and will therefore exit the decision tree at Q1. But don’t make assumptions here!

Answering ‘no’ to Q1…
If you answer ‘no’ then you progress to the next question.

Question 2

‘Do specific control measures for the identified significant hazard exist at this step?’

Q2 and Q4 both focus on the specificity of the control measure/step to control the significant hazard. At the time of writing, I have some concerns around Q2 (but we do need to cogitate on these things!)

Answering ‘no’ to Q2…
If you answer ‘no’ to this question, the new version states ‘this step is not a CCP and subsequent steps should be evaluated for a CCP’ which I do agree with as controls around a CCP must be specific. However, if we’ve said ‘no’ to Q1 the significant hazard isn’t controlled by PRPs and we’ve said in Q2 that we don’t have specific control measures for the identified significant hazard at this step then what/where are the control measures?! This would red flag to me! There is a footnote to the ‘no’ answer on this question, which states ‘if a CCP is not identified at questions 2-4, the process or product should be modified to implement a control measure’ – this is effectively like asking the previous Q1a ‘is control necessary at this step for safety’. Are you still with me?!

Answering ‘yes’ to Q2…
If you answer ‘yes’ then you progress to the next question.

Question 3

‘Will a subsequent step prevent or eliminate the identified significant hazard or reduce it to an acceptable level?’

Q3 bring us back to more familiar territory. You will recognise this as being Q4 in the previous version. The wording has been expanded slightly to include the word ‘prevent’ which I think is a good addition as some people saw this as implied in the previous version, whilst others were sticklers on the omission of the word ‘prevent’.

Answering ‘yes’ to Q3…
If you answer ‘yes’ to Q3 then the decision tells you that the subsequent step should be a CCP, (the previous version stopped short of making that decision for you). However, I do think this subsequent step should still be carefully considered to avoid errors or assumptions. Q3 includes a footnote on both the question and ‘no’ answer. The footnote on the question draws attention to whether the control measure at this step works in combination with a control measure at another step to control the same hazard (such as a multi-stage heating or cooling process), in which case the tree tells us that both steps should be CCPs – again we need to keep a good dollop of logic here and remember that the tree is a tool for decision making and not the be all and end all!

Answering ‘no’ to Q3…
If you answer ‘no’ to Q3 then you progress to Q4.

Question 4

‘Can this step specifically prevent or eliminate the identified significant hazard or reduce it to an acceptable level?’

Q4 is similar to Q2 in the previous version.

Answering ‘yes’ to Q4…
If you answer ‘yes’ to Q4 this takes you to the ‘this step is a CCP’ conclusion. Note that there is now only one route to the ‘this step is a CCP box’, in the previous version there was a direct route from answering ‘yes’ to Q2 (is the step specifically designed to eliminate the likely occurrence or reduce it to an acceptable level).

Answering ‘no’ to Q4…
If you answer ‘no’ to Q4 you are instructed to modify the step, process or product to implement a control measure. I think we need to keep in mind that the wording of the new version Q4 is ‘can the step specifically prevent or eliminate…’ whereas in the old version Q2 was ‘is the step specifically designed to… I think there is a difference between ‘can’ and ‘is’.

A couple of examples:

decision tree examples from codex 2023


In conclusion, this is a good step forward for Codex and brings it more in line with modern day HACCP management.

If you’d like to increase or refresh your HACCP skills join one of our HACCP courses, or test your knowledge on our Level 3 HACCP or Level 4 HACCP quizzes.


  1. Abdul Azeez says:

    The quostion number 4 is well designed to ensure that the right ccp is identified . Some cases I have seen wrong ccp is identified .

    • Jessica Murray says:

      Thanks for your comment. Yes I agree. I think the new version will help to reduce the number of potentially incorrect CCPs.
      – Adele

  2. Victor Muliyil says:

    This decision tree is inaccurate in two places. I have tried to contact Codex & the CFIA over 20 years ago to explain this to them but I got no response.

Leave a Reply

Your email address will not be published. Required fields are marked *