RESIDUES AND EXPORT

Codex Revoked Every MRL for Five Pesticides: What South African Exporters Must Check Now

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

At CAC48, the 48th Session of the Codex Alimentarius Commission held from 10 to 14 November 2025 in Rome, all Codex maximum residue limits were revoked for fenthion, parathion-methyl (also written methyl parathion), dinocap, amitraz and bitertanol, and methamidophos MRLs were revoked for cottonseed, fodder beet, potato and sugar, while 340 maximum residue limits for different combinations of pesticide and commodity were adopted, a figure confirmed in the adopted report REP25/CAC.

What is a Codex MRL, and what is it not

A Codex maximum residue limit is the highest concentration of a pesticide residue recommended by the Codex Alimentarius Commission to be legally permitted in or on a named food or feed commodity. It is a voluntary international reference standard, expressed in milligrams per kilogram. It is not law anywhere until a country adopts it, and Codex cannot ban, register or withdraw a pesticide.

That distinction costs exporters money. An MRL is a good agricultural practice figure that has passed a health check: if the product is used correctly, at the approved rate and with the approved pre-harvest interval, the residue should not exceed this figure, and the dietary intake that follows was assessed before the limit was recommended. A result above the MRL is first evidence that the product was not used to the approved good agricultural practice, and only after that a toxicological question. For most actives that makes a detection a trade compliance failure rather than an immediate health event.

It is not always so. Where the active carries an acute reference dose, and the organophosphates on this revocation list do, an exceedance on a commodity eaten without further processing has to be assessed against acute dietary exposure before you decide whether product already on the market can stay there. Do that assessment before you treat a detection as a border problem.

The Commission adopts what its committees have forwarded once a year, in one batch. That annual rhythm is why a whole block of MRLs can appear or disappear on a single date, and why a specification written eighteen months ago can be out of date without anybody at your site touching it.

Limits are set on the back of supervised residue trials and a dietary exposure assessment, evaluated jointly by FAO and WHO scientific experts before a committee forwards a recommendation to the Commission. That process is slow and it is data hungry, which is the practical reason limits get withdrawn: when nobody is left to supply current data for an old active, the limit eventually goes rather than sitting on the list unsupported.

An MRL is a good agricultural practice number that has passed a health check, not a health limit derived from one. It says the product was used the way the label said, and that the resulting dietary intake was assessed against the ADI and the acute reference dose before the limit was recommended.

A Codex MRL is also not a guarantee that your buyer will accept the residue, because private specifications routinely sit below the legal limit, and it is not a substitute for the destination country’s own rule, which always wins where one exists. We set out the relationship between Codex, the WTO and South African law in is Codex Alimentarius law in South Africa, and the full 2025 and 2026 picture in the full list of every Codex change from CAC48 and CAC49.

Which pesticide MRLs Codex revoked in November 2025

At CAC48, held from 10 to 14 November 2025 at FAO headquarters in Rome, all Codex MRLs were revoked for five actives: fenthion, parathion-methyl, dinocap, amitraz and bitertanol. At the same session 340 maximum residue limits were adopted for different combinations of pesticide and commodity, a figure confirmed in the adopted report REP25/CAC.

Notice the shape of that list. These are older actives, several already withdrawn or heavily restricted in the markets that matter most to South African fruit. A revocation of this kind usually means the supporting data is no longer current or the active is no longer supported by a registrant. The regulatory logic is housekeeping. The commercial consequence is not.

How the Codex fallback actually works in trade

Where an importing country has not set its own MRL for an active on a commodity, its import rules generally do one of three things: fall back to the Codex limit, allow an import tolerance, or apply a default limit at or near the limit of determination. Remove the Codex MRL and the first route closes. A consignment then faces a default that is effectively zero.

This is the part that strands containers. A default at the limit of determination is not a lenient number. It means the analytical method decides your compliance, and as laboratory sensitivity improves the effective limit falls without anybody legislating anything. A residue that was invisible and therefore compliant five years ago is detectable and therefore a violation today, on the same field, with the same spray programme.

South Africa exports a very large volume of fresh produce, and fresh produce is where this hurts most, because you cannot rework it and the clock on a citrus or table grape consignment runs in days. A detention on residue grounds is not a laboratory dispute. It is demurrage, a re-inspection, a possible destruction or re-export order, a claim from the buyer, and a listing on that authority’s notification system that follows your exporter code around next season.

An import tolerance is worth understanding properly, because exporters talk about it as if it were a form to complete. It is an application, usually supported by residue trial data and a dietary exposure assessment, made to the importing authority by a registrant, an exporting government or an industry body, and it takes time and money that a single season does not allow. If your market access depends on one being granted, that is a project to start now with your industry body, not a fix available to a pack house in January.

• The Codex MRL disappears, but the active can still be legally registered and applied somewhere in your supply chain
• The destination’s own MRL, where one exists, is unaffected and still applies
• Where the destination had no MRL and relied on Codex, there may now be no permitted level at all
• An import tolerance is applied for, granted and published. It is not something you can assume
• Your buyer’s private specification may already sit below whatever the legal limit turns out to be
• Your own residue test result is only meaningful once you know which limit you tested against

Why importing authorities lean on Codex figures

Under the WTO framework, Codex is the named international reference for food safety measures, which is why importing authorities lean on Codex figures. An authority that sets its limit at the Codex figure is working from the recognised international reference, and another may set a different limit of its own. Confirm the position for your destination market before you rely on it.

That reliance is why Codex matters commercially to a business that never reads a Codex text. It gives an importing authority a defensible position at almost no cost, and it explains why so many countries write a Codex fallback into their residue rules rather than setting thousands of national limits for actives they may never see. Follow it through to a revocation and the authority is left to set its own limit, apply a default, or handle import tolerance applications one at a time. None of those is fast, and two of them are hostile to a consignment already on the water. The mechanics are set out in our companion article on whether Codex is law in South Africa, which is the piece to read before you argue this with a customer.

What a South African exporter must check this month

Four steps, in this order, before next season’s programmes are locked in. List every active used on the crops you grow or buy. Check each one against the destination market’s own published limits. Identify where you or your customer were relying on a Codex limit rather than a national one. Then get the position in writing from the party who will be held to it.

1. List the actives, by crop and by blockNot the products, the actives. A grower buys branded products and an authority tests for actives, and the two lists are never the same length. Build it from spray records for the full season, including post harvest, in store and in transit treatments, and anything applied by a contractor.
2. Match each active to every destination you ship toOne consignment can go to several markets, and one market can hold several limits for the same active across commodities. Build a matrix of active against destination against commodity. The empty cells are the risk, because an empty cell means no national limit, which is exactly where the Codex fallback used to sit.
3. Flag the five revoked actives and the four methamidophos commoditiesFenthion, parathion-methyl, dinocap, amitraz and bitertanol on any crop, and methamidophos on cottonseed, fodder beet, potato and sugar. If any appear in your spray records, in a supplier declaration or in a stored product treatment, that line needs an answer before the season.
4. Get the destination position in writingFrom the importing authority, the buyer’s technical department or your certification body, and dated. A verbal assurance from an agent is worth nothing when a consignment is detained. File it against the raw material specification so the next person to open the file sees the same answer you did.
5. Decide what changes in the spray programmeIf an active has no lawful destination limit, the practical answer is usually to remove it from the programme for export blocks and to say so in the grower agreement, with an alternative agreed. That decision belongs to an agronomist and a technical manager together, before planting.

Name an owner for that exercise before you start it. In most pack houses it falls between the technical manager, who knows the specifications, the agronomist or grower liaison, who knows the programmes, and the export or logistics manager, who knows the destinations. Left unassigned it does not happen, and the first person to discover the gap is whoever takes the call about a detained container.

Do the same for feed. Fodder beet and cottonseed are on the methamidophos list, and a feed input carrying a residue can reach a food of animal origin later. If you supply a dairy, an egg producer or a red meat abattoir, that chain is yours to assess.

Grower declarations and spray records for a residue claim

A revoked MRL turns into a rejected consignment through your supplier approval process, which is where it should be caught. Approval of a grower or a pack house has to reach past the certificate to the plant protection products actually used: a signed declaration of actives, dates, rates and pre-harvest intervals, spray records traceable to the block, and a change notification undertaking that reaches you before the fruit does.

Most South African pack houses already collect a grower declaration. The weakness is rarely that the document is missing. It is that the declaration lists product trade names rather than actives, that nobody reconciles it against the spray records, and that it is signed once in October and never revisited when the programme changes in February under a mildew pressure nobody predicted.

Score the grower and the material together. A well run grower supplying a heavily sprayed crop into a market with a default limit is a higher risk than a weaker grower supplying a low input crop into a market with generous published limits. That combined scoring is what the Supplier and Purchased Material Approval Risk Assessment Template (RA11) at R690 is built to do, and the Chemical Contamination Risk Assessment Template (RA16) at R690 carries the hazard side of the same question. To build the capability in house, start with our guide to building internal and supplier auditors.

The two residue clauses your raw material specification needs

A raw material specification that says “complies with applicable legislation” has said nothing, and it will not help you when a consignment is detained. The specification should name the destination markets, the limit applied for each relevant active, the source of that limit, the date it was checked and who is responsible for re-checking it. That last line is the one almost nobody writes.

Put a review trigger on the specification rather than a review date, because Codex adopts in an annual batch and your annual review may fall four months after the change. Better triggers are a new Codex batch, a change to the destination market’s published limits, a change of grower or growing region, a change in the spray programme, and a customer specification revision. Say plainly on the specification where a limit is uncertain: a line recording “no national limit identified for this active in this market as at this date, confirmation requested from the buyer” is defensible, while a limit copied from a Codex table that no longer exists is worse than nothing, because someone downstream will rely on it.

Customer specifications deserve the same treatment. A retail programme that works to a fraction of the legal limit has set your operating target, and where the legal limit has been revoked that fraction has nothing to be a fraction of. Ask the customer to state the absolute figure they will apply, in milligrams per kilogram, rather than a percentage of a number that may no longer exist, and record their answer on your own specification with the date.

How much residue testing you need, and on what basis

Test on a risk basis, not on a fixed number per season pulled out of the air. The variables that should set frequency are the crop and its typical spray load, the grower’s history and control level, the destination market’s limit type, the actives in the programme, the season’s pest pressure, and any history of detection on that grower or commodity. Write the basis down.

Two practical points. A multi residue screen does not cover every active, and the ones it misses are often the older actives needing a single residue method, so ask the laboratory for its scope and check your declared actives against it. And agree what limit each result will be reported against, printed on the report, because a result reported as “not detected” with no reporting limit is not a usable document in a dispute.

• Set frequency from crop, grower, destination and active, and record the reasoning on the plan
• Check the laboratory’s multi residue scope against your actual declared actives, active by active
• Have the reporting limit and the limit tested against printed on every certificate
• Use an accredited laboratory and hold the scope of accreditation on file, not just the logo
• Trend results by grower and block instead of filing them, so a drift is visible before a failure
• Define what happens on an out of specification result before you get one, including who holds the consignment

CXG 106-2025 and the expired reference material your laboratory can still use

CXG 106-2025, Guidelines for Monitoring the Stability and Purity of Reference Materials and Related Stock Solutions of Pesticides During Prolonged Storage, was adopted at CAC48 in November 2025. It lets a laboratory keep using a pesticide reference material past its stated expiry date where purity is demonstrated to remain acceptable. For a residue laboratory holding hundreds of certified standards, that is a direct and recurring cost saving.

Reference materials for pesticide analysis are expensive, they arrive with conservative expiry dates, and a laboratory running a broad multi residue scope replaces them constantly. Under this guideline an expiry date becomes a prompt to verify rather than an instruction to discard, provided the monitoring is documented and can be shown to an assessor. If you are a food business rather than a laboratory, it gives you a sensible question to ask a laboratory tendering for your work: how do you monitor stability and purity during prolonged storage, and may I see it. The text is a free download, as every Codex text is: CXG 106-2025 on the FAO Codex site, indexed in the ASC Codex and guidelines register alongside 322 Codex texts with the full 2025 and 2026 change log.

CXM 2, camelid MRLs and coccidiostat action levels in chicken eggs

At CAC49, the 49th Session held from 6 to 10 July 2026 in Geneva, CXM 2, Maximum Residue Limits and Risk Management Recommendations for Residues of Veterinary Drugs in Foods, gained the first camelid MRLs, covering tetracyclines in camelid kidney, liver, milk and muscle and ivermectin in camelid milk, together with action levels for lasalocid and nicarbazin in chicken eggs.

The published CXM 2 text has not caught up yet, so confirm before you specify.

One caution applies here. The published CXM 2 PDF still reads CXM 2-2024 and carries none of these changes, so confirm the position against the Codex text or send it to ASC to confirm before you write it into a specification or a customer commitment.

For a South African egg producer the coccidiostat point is the one to sit with. Lasalocid and nicarbazin are used in broiler feed, and the practical risk to a layer flock is carry over: feed intended for broilers reaching layers through a shared mill, a shared auger, a shared bulk bin or a delivery sequence nobody flushed. An action level is not an MRL. It is a level at which a competent authority is expected to act, and it exists precisely because unavoidable carry over is a known problem in feed manufacture.

• Ask your feed supplier in writing whether medicated feed and layer feed share lines, augers, bins or delivery vehicles
• Ask what flushing sequence runs between medicated and non medicated batches, and ask to see the record
• Ask for the supplier’s own carry over testing data rather than a statement of policy
• Put coccidiostat carry over on the raw material risk assessment for feed, with a defined verification

The full text sits at CXM 2 on the FAO Codex site. Note while you are there that the July 2026 session was CAC49 and not CAC48. Calling the July 2026 session CAC48 is the most common error in circulation, and it matters here because the pesticide revocations belong to the November 2025 session and the veterinary drug changes to the July 2026 one.

How to check whether the South African MRL differs from the Codex figure

Codex is not South African law. South Africa sets its own residue limits, and this article deliberately publishes no South African MRL figure and names no South African regulation number, because ASC has not verified them for this piece and a wrong number would be worse than none. Check the ASC legislation register, 413 instruments across 42 parent Acts, or ask us to confirm the position for your commodity.

Start from the South African pesticide residue regulation and its amendments.

That restraint is the standard we hold ourselves to. Residue regulation is amended by notice, the amendments are commodity specific, and a figure quoted from memory or from a secondary source is how a site ends up defending a specification it cannot support. If you need the current South African limit for a specific active on a specific commodity, that is a question with a documented answer, and one we answer for clients as part of the labelling and regulatory advisory service.

Start with the ASC South African legislation register for the instruments themselves, and read our South African food legislation guide for how they fit together. If you import as well as export, importing food into South Africa covers the inbound side, where Port Health inspects imported foodstuffs at points of entry and can detain a consignment.

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.