LISTERIA

Listeria After the 2026 Codex Revision: The Wider Ready to Eat Definition and Your Environmental Monitoring Programme

By Mthokozisi Nkosi, Food Safety Specialist & Lead Auditor, ASC Food Safety · 14 min read

CXG 61-2007 was fully revised at CAC49 in July 2026 and the published text carries REVISED 2026. The revision builds the ready to eat definition on what consumption is reasonably foreseeable, and Codex publishes no list of the foods it reaches, so you apply it to your own range. A product that sat outside scope last year may sit inside it now, which re-scopes the hazard analysis and the environmental monitoring programme with it.

What the 2026 revision of CXG 61-2007 actually changed

CXG 61-2007, the Guidelines on the Application of General Principles of Food Hygiene to the Control of Listeria monocytogenes in Foods, was fully revised at CAC49 in July 2026 and the published text carries REVISED 2026. The headline change is the ready to eat definition, which now turns on what consumption is reasonably foreseeable. Codex publishes no list of the foods it reaches, so you apply the definition to your own range.

The revised definition reads: “Any food (raw or processed) for which it is normally eaten without further validated treatment sufficient to achieve food safety, or for which it is reasonably foreseeable, based on evidence of consumer habits or practices, that it will be eaten without such treatment.” The second half is the new part, and it is the half that moves products into scope.

A full revision is a different event from an amendment. An amendment edits named sections and leaves the architecture in place. A full revision reopens the whole document, which is why the July 2026 text is worth reading end to end rather than skimming for a change bar. For most South African manufacturers, though, the consequence lands in one place: the boundary of the word that decides whether the rest of the guideline applies to your product at all.

Get the session right, because most of the summaries circulating have it wrong. The revision was adopted at CAC49, the 49th Session of the Codex Alimentarius Commission, held from 6 to 10 July 2026 at the Centre International de Conferences Geneve in Geneva, with about 600 delegates from 119 members present. CAC48 was the November 2025 session in Rome. If a supplier newsletter or a consultant’s briefing tells you the Listeria guideline changed at CAC48, that briefing was not written from the source, and you should treat the rest of its detail with the same suspicion.

What reasonably foreseeable consumption means in practice

Reasonably foreseeable consumption moves the test away from what your label instructs and towards what a normal consumer can be expected to do with the product. A cooking instruction does not remove a food from scope if a meaningful share of buyers will eat it without applying that step. The question is not what you intended. It is what happens in real kitchens.

That is a harder test to argue against than the one most sites have been using. The old habit was to read the pack: if the label said cook before eating, the product was not ready to eat, the Listeria hazard was controlled by the consumer’s own cook step, and the hazard analysis said so in a single line. The new reading asks whether the consumer’s step is a control you are entitled to rely on. Where the answer is no, the lethality has to come from somewhere inside your factory, or the product has to be managed as ready to eat.

Two practical points follow. First, foreseeable is not the same as possible. Somebody could eat raw mince, and that does not put raw mince inside the definition. The test is whether consumption without a lethality step is a pattern of use rather than an outlier. Second, foreseeable is evidence based and it moves. Consumer habits change, marketing changes them faster, and a product that was genuinely always cooked five years ago may not be today. Your own recipe suggestions, your social media, your packaging photography and your category’s usage research are all evidence, and an auditor is entitled to read them alongside your hazard analysis.

Why diced cantaloupe, frozen vegetables and cold smoked fish are where the foreseeable test bites hardest

The revised CXG 61-2007 names these three in its Introduction as the foods for which JEMRA has published open source quantitative risk assessment models, MRA 38, 47 and 48. Codex publishes no list of in scope foods. They are useful here because each fails a different comfortable assumption: that an intact skin protects the flesh, that freezing is a control, and that a traditional process is a lethality step.

Diced cantaloupe. Whole melon carries a rind that is grown in soil, handled repeatedly and rarely washed by the consumer with any conviction. The moment the fruit is cut, the knife moves whatever is on that rind into flesh that has a near neutral pH and a high water activity, and the product is then held chilled for days. There is no kill step anywhere in that chain, and nobody cooks a fruit salad. The hazard is created by the process the manufacturer performs.

Frozen vegetables. Freezing does not kill Listeria monocytogenes. It injures a proportion of the population and lets the rest decline very slowly, far too slowly to count as a reduction, and the injured cells recover once the product thaws. Blanching before freezing is designed for enzyme inactivation and colour, and it is almost never validated as a lethality step for the pack that leaves your site. Meanwhile the consumer adds frozen peas and spinach to smoothies, defrosts sweetcorn straight into a salad, and picks at frozen berries from the bag. Codex observes that some foods linked to listeriosis outbreaks, certain frozen vegetables among them, were intended by the food business operator to be cooked by the consumer, which is exactly where the foreseeable test bites.

Cold smoked fish. Cold smoking is a flavour and preservation process that runs deliberately below the temperatures that would cook the flesh. It is sliced, packed, held chilled and eaten as it comes. This is the classic example, and the site that makes it usually knows. What the revision does is remove the argument that similar products, made by a similarly gentle process, are somehow different because nobody has said so out loud.

A scope test you can run across your own range this week

The revised guideline gives you the principle. It does not hand you a worksheet for your factory. The test below is the sequence ASC uses on site to decide whether a product falls inside the ready to eat definition, and it takes an afternoon for a range of thirty lines if the technical file is in order and the marketing material is to hand.

1. Name the last validated lethality stepNot the last hot step. The last step you have validation data for, applied to the pack that leaves your site. If there is none, write none, and stop pretending the blancher counts.
2. Ask what happens to the product after itIs product open to the room after lethality, and for how long? Slicing, dicing, portioning, weighing, hand packing and cooling in an open tray are all post lethality exposure.
3. Describe how the product is really eatenRead your own pack, website and campaigns. If any of them shows the product eaten without heat, foreseeability is settled and there is nothing left to argue.
4. Ask whether the consumer step is a control you may rely onA control has a specification, a measurement and a record. A consumer’s frying pan has none of the three. Relying on it needs a stated justification, not a habit.
5. Check whether the product supports growthWater activity, pH, preservatives, competing flora and the storage temperature on the label. This decides how hard the next steps have to work, not whether the product is in scope.
6. Consider who eats itFoods reaching pregnant women, older consumers, infants and immunocompromised people carry the consequence that makes this organism a listeriosis problem rather than a spoilage problem.
7. Write the decision down either wayAn out of scope decision is a decision. Record the product, the date, the reasoning, the evidence you read and the author. An undocumented out of scope call reads at audit as an oversight.

The finding is rarely that a product is in scope. The finding is that nobody ever asked, and there is no record of the question.

Does the Codex revision change South African law?

No. Codex is not South African law. R638 of 2018, the Regulations Governing General Hygiene Requirements for Food Premises, the Transport of Food and Related Matters, is unchanged and names no environmental monitoring programme. The revision reaches a South African manufacturer through certification schemes, export customers and specifications, and as the reference an expert will use where our own regulations are silent.

That last route is the one people underestimate. Where a South African regulation says nothing, the question of what a reasonable manufacturer should have done gets answered from somewhere, and Codex is the most likely somewhere. Section 61 of the Consumer Protection Act 68 of 2008, “Liability for damage caused by goods”, imposes liability on producers, importers, distributors and retailers for harm caused by unsafe goods, product failure, defects or hazards, or by inadequate instructions or warnings, irrespective of negligence. A file that shows you asked the scope question in September 2026 and recorded the answer is worth a great deal in that setting. A file that shows nothing is worth what it looks like.

This organism is taken seriously in South Africa, by regulators, by retailers and by technical teams. That is the context every South African retailer specification for chilled ready to eat product is written in, and it is why a wider Codex definition will be tested here through customer audits sooner than through any change in our own regulations. Our South African food legislation guide sets out the wider legal duties this sits inside, and the full list of every Codex change from CAC48 and CAC49 covers everything else Codex adopted in July 2026.

What a newly in scope product needs before your next audit

A product that moves into the ready to eat definition needs three things that it probably does not have: a Listeria hazard analysis that treats the organism as credible, a shelf life position that states whether the organism can grow across the declared life, and an environmental monitoring programme that actually looks for it on the line that makes it. Everything else follows from those three.

The reason the change bites is that these three documents are usually written once, early, when the product was classified as something else. The hazard analysis says controlled by consumer cooking. The shelf life was set on spoilage and sensory data, because nobody was asking a pathogen question. The monitoring plan covers the chilled ready to eat line at the other end of the factory and never mentions the freezing tunnel or the dicing table. None of the three is wrong on its own terms. All three are now answering a question that has changed.

How to write the Listeria hazard analysis for a product that has just come into scope

Reopen the hazard analysis at the process step, not at the product. Identify the last validated lethality step, then every step after it where product is open to the room, and treat each one as a point where recontamination is credible. The line that used to read controlled by consumer cooking becomes a set of hygiene controls you own, with verification attached to each of them.

The severity score is the part sites get wrong. Listeriosis reaches comparatively few people and is severe in those it does reach, which is a different shape of risk from most foodborne pathogens. Scoring severity as moderate because positives are rare misreads the organism. Score likelihood on your own factory and your own data, and score severity on what the illness does to a pregnant woman, a newborn or an immunocompromised adult.

Then be honest about what your controls are. Cleaning and disinfection is a control, and it needs a frequency, a chemical, a contact time and validation behind it. Zoning is a control, and it needs a drawn map. Air handling, drainage falls, equipment design and traffic flow are controls. Environmental monitoring is verification that the controls are working, and the value of a negative result depends entirely on whether the plan was capable of producing a positive one. The design method for that programme is set out in full in our guide to microbiological risk assessment and environmental monitoring, and this article deals only with what a scope change does to it. If your HACCP plan itself needs rebuilding around the new scope, start from what a HACCP plan has to contain.

Shelf life: growth or no growth, and where the challenge study sits

A ready to eat product with a credible Listeria hazard needs a shelf life position that stands up. The question is narrow: across the declared shelf life, under the storage conditions on the label, does Listeria monocytogenes increase by more than 0.5 log or not? That 0.5 log is the working convention challenge study laboratories and retailer specifications use.

A no growth product is managed with a contamination argument. A growth product is managed with a contamination argument and an evidenced time limit.

Most sites set shelf life on spoilage and sensory endpoints, which answers a different question. A product can look, smell and taste correct on day fourteen and carry a population that has increased by orders of magnitude since day one, because this organism grows without spoiling the food and grows at refrigeration temperature. That is the whole difficulty with it.

A challenge study inoculates your product with the organism, stores it under conditions that represent real distribution including a realistic temperature abuse leg, and follows the population across and beyond the declared life. It is the strongest evidence available, and it is also the study most often designed badly: too few time points, one batch, an unrealistically cold storage regime, a strain cocktail chosen for convenience, and no account of what happens when a retailer’s display cabinet runs warm. Design it before you commission it, because a laboratory will run what you ask for.

Where the formulation clearly prevents growth, documented intrinsic factors and predictive modelling may carry the position without a challenge study. Where the product supports growth, or sits close to a limit, the study is usually the only thing that will satisfy a retailer’s technical team. Either way, write the conclusion as a sentence a non specialist can read, and file the data behind it.

What to add to an existing swab plan when a product comes into Listeria scope

Re-scoping is not the same as designing from scratch. The zone model, the organism logic and the response protocols you already run for your chilled ready to eat line are the method. What a scope change does is extend that method over an area of the factory it never covered, and the work is mostly map, sites, frequency and the rules that come with them.

Start with the zone map, because a scope change usually breaks it. An area that made a raw or cook before eating product was zoned as a raw or low care area, with the traffic, protective clothing and cleaning standards that follow. If a product from that area is now ready to eat, either the area moves up in the hygiene zoning or the product’s exposure after lethality has to be closed off. That is a layout argument before it is a monitoring argument, and the two documents have to agree.

• Redraw the zone boundary for any area now producing a ready to eat product, and show why the boundary sits where it does
• Add the monitoring sites the old plan never had: the dicing table, the freezing tunnel, the slicer, the packing conveyor, the post lethality cooling area
• Plot every drain in the newly in scope area, with the direction of fall, because drains connect zones you assumed were separate
• Include the equipment nobody swabs: hollow rollers, worn belt edges, cracked conveyor seals, chipped cutting surfaces, the underside of the machine
• State whether each new site is sampled during production or after cleaning, and keep the two apart in the record
• Raise the frequency on the newly in scope line for an initial baseline period, then set the routine frequency from what you actually find
• Extend the response protocol to the new sites before the first result comes back, not after

Frequency deserves a specific decision rather than a copied one. A line with no Listeria history has no baseline, so the first few months are not routine verification, they are characterisation. Sample it harder than you intend to sample it in a year, expect to find something, and set the steady state frequency once you know what the line does. Sites that start a newly in scope line on the same low frequency as a mature, well characterised line learn nothing for a year and then get a surprise.

Which mode of monitoring a scope change calls for

Routine verification samples the same defined sites on a defined frequency to show the controls are holding. Seek and destroy goes looking for harbourage: equipment stripped down, hollow frames opened, worn surfaces sampled, maintenance in the room. A scope change calls for seek and destroy first. Routine verification of an area you have never characterised verifies nothing, because you do not yet know where to look.

Run the first sweep with production, engineering and hygiene in the room together and with the equipment apart. The point is to sample the places that never get sampled: inside hollow rollers and frames, under worn belt edges, behind cracked welds, inside cabinet bases, under floor level equipment feet, in the drainage that runs beneath the newly in scope line. Take the samples during or immediately after production rather than after a full clean, because a cleaned surface tells you about the clean and not about the harbourage underneath it.

Budget for what you find. A seek and destroy event that identifies a persistent site inside a piece of equipment usually ends in a repair, a replacement or a design change, not in a stronger chemical. If the outcome of every sweep at your site is a deeper clean, the sweeps are not being read honestly. Approach it expecting to spend money, and the exercise pays for itself the first time it prevents a product decision.

What a positive means on a newly in scope line

Write the response protocol before the first sample leaves the building, and write it for two organisms. Most programmes screen for Listeria species, because the genus is the wider and more sensitive net. What a positive triggers depends on which of the two it is, and on where the swab was taken.

A Listeria species positive on a product contact surface where ready to eat product is exposed after lethality triggers speciation, intensified sampling around the site, and a hold on the affected batches while the speciation runs. A confirmed Listeria monocytogenes positive on that same surface is a product decision in its own right, and holding is the sensible default while you establish what ran on that surface. Positives on remote processing surfaces and outside processing are investigation triggers at either level rather than automatic product decisions.

The sequence that holds up is the same one you already use elsewhere. Contain, by identifying every batch that ran on or near the site since the last clean result. Notify the named person who is allowed to make the product decision, inside a stated time. Investigate to root cause, which means intensified sampling around the site rather than a single re-swab. Correct, and then verify the correction with a sampling intensity that would have found the original problem. Close the record with what changed, not with the phrase area cleaned.

Then use the trend. A trend is real when it changes a site, a frequency, a cleaning method, a piece of equipment or a traffic route, and it is decoration when it produces a graph in the management review pack that nobody acts on. For a newly in scope line, the trend has a specific job in the first year: it tells you whether your initial frequency and site selection were right, and it should be reviewed on a shorter cycle than your established lines while that question is open.

What else changed at CAC49, and what did not

CXG 78-2011, the Guidelines for the Control of Campylobacter and Salmonella in Chicken Meat, was fully revised at the same session, its first revision since 2011, incorporating JEMRA scientific advice and adding coverage of antimicrobial resistance and storage temperatures. That was adopted. The consolidated text had not been republished by FAO when ASC checked on 10 September 2026.

The practical instruction that follows is narrow. You may plan on the basis that the revision was adopted, and you should confirm the wording against the published guideline before you write any of it into a specification, a supplier requirement or a customer commitment. Poultry processors and the sites buying from them will want the storage temperature provisions in particular, and those are exactly the sort of provision that should not be quoted from a session summary. The ASC Codex and guidelines register carries 322 Codex texts and 60 other guidance documents with the full 2025 and 2026 change log, and it is updated as the consolidated texts appear. Every Codex text is a free download from FAO, so nobody should be selling you one.

What did not change matters just as much. CXC 1-1969, the General Principles of Food Hygiene, did not change in 2026. The published text still reads Adopted 1969. Amended 1999. Revised 1997, 2003, 2020, 2022. Editorial corrections 2011. There is no 2026 edition, whatever you may have read. The 2022 revision, with its HACCP annex, remains the current text, and it remains the prerequisite programme and HACCP baseline that every Listeria control in this article sits on. A revision is in progress at the Committee on Food Hygiene, but it has not been adopted, and until it is, your prerequisite programmes are measured against the 2022 text.

For a certified site, the scheme is where all of this becomes an audit question. BRCGS Global Standard Food Safety Issue 9 carries clause 4.11 as one of its twelve fundamental requirements, and FSSC 22000 Version 7 carries environmental monitoring in its additional requirements. Neither scheme will ask whether you have read the revised Codex guideline. Both will ask why a ready to eat product is not covered by your programme, and reasonably foreseeable consumption is now the reference the answer is measured against. Our guides to implementing BRCGS Issue 9 and the FSSC 22000 Version 7 guide for South Africa set out how each scheme handles it.

Read next from the ASC library

Frequently asked questions

Published by ASC Food Safety, South African food safety and quality consultants. This article is general guidance and not a substitute for certification-specific advice.