The Jar of Baby Porridge

On 5 and 6 May 2026, the European Commission's standing committee on pesticides had two papers in front of it.

One would extend the approvals of twenty-one substances. Among them were fludioxonil, a fungicide used on fruit across Europe, and cyprodinil, the fungicide it is usually mixed with.

The other would end cyprodinil's approval.

Neither was put to a vote.¹

Both came back to the agenda on 29 June. No summary of that meeting has been published.

Fludioxonil's approval runs out on 30 September 2026.²

Fludioxonil meets the international definition of a PFAS. Cyprodinil contains no fluorine, and does not. The line between them is a real chemical line.

The only apple in the file

Search Britain's 2024 pesticide residue file for an apple and you find one. It is a jar of oat and apple porridge for infants.

No residue was detected in it.³

That jar is the only apple in the file because the monitoring programme did not test apples that year. It did not test them in 2023 either. Pears were surveyed; apples were not.

The programme rotates. In 2024 it covered grapefruit, melon, grapes, bananas, limes and about twenty other foods. In 2023 it covered pears, oranges, kiwi fruit, carrots and a similar spread.

The file itself is remarkable.

For every sample it gives the product, the country it came from, the retailer, the retailer's store address and the packer. Then every residue found, its level, and the legal limit.³

A determined person can establish which supermarket sold a particular grapefruit, which street that shop is on, and how much fungicide was on the fruit.

Count the words in the file and you get this:

• fludioxonil — 160
• cyprodinil — 75
• PFAS — 0³

Somebody already did this

Strawberries were not in the 2023 or 2024 programme. They were in the 2022 one.

In April 2024, Pesticide Action Network UK — a campaigning organisation — went through the government's own results and marked which pesticides meet the PFAS definition. Its figures cover that whole class. They are not about fludioxonil, which PAN UK never singles out.

It found ten different PFAS pesticides across UK fruit, vegetables and spices. On strawberries, PAN UK wrote:

"Strawberries were found to be the worst offenders, with 95% of the 120 samples tested by the government in 2022 containing PFAS pesticides."⁴

That covers pesticides anywhere in the class, not fludioxonil specifically.

PAN UK also counted the UK-grown samples separately. Of 55, it found 50 carrying residues of pesticides in the same class — 91%.⁴

The sampling is the government's. The sorting is PAN UK's. The sample counts here are PAN UK's reading of the 2022 dataset rather than a fresh count of it.

So the connection has been made. Publicly, two years ago, by a named organisation, using the government's own numbers.

PAN UK also found something further back down the chain:

"Farmers are generally unaware that they are spraying 'forever chemicals' on their crops because there is no information on the label."⁴

The person applying it is not told either.

Defra tells you to go shopping

The Department for Environment, Food and Rural Affairs put up a blog post for the general public on 3 February 2026. That was twenty-two months after PAN UK's analysis. It is called "Forever chemicals: the problem and our plan."⁵

There is a stock photograph of a family walking in the countryside. A section heading reads "Here's what you need to know."

The post explains the word. PFAS are "a family of thousands of man-made chemicals," brilliant at resisting heat, water and oil, and once they are loose they "don't break down naturally."

Then it says where they are. Stain-resistant carpets. Waterproof jackets. Some cosmetics. Food packaging, like takeaway containers.

Then it tells you what to do about it.

"If you want to reduce your exposure you can look for PFAS-free alternatives when shopping."⁵

The words fruit, vegetable, crop, residue and fungicide do not appear anywhere on that page. Pesticide appears once, in the blog's own category tag, and never in the body. The only food it mentions is the packaging around the food.

Two plans that never meet

A blog is a blog. The strategy documents behind it are the better test, and there are two of them.

The UK PFAS Plan, published in 2026, mentions food 28 times. It has a section headed "Managing the risks from PFAS in the food we eat." It describes validated testing for PFAS in fish, shellfish, beef, poultry, milk and eggs, with work ongoing for fruits, vegetables and cereals.⁶

Saying that nobody in government connects PFAS to food would be false. That workstream exists.

What is not in it is the route. Across the whole plan, residue does not appear. Neither does crop, nor plant protection, nor fungicide.

Pesticide appears once. The sentence it appears in is about a different chemical, and it is about rivers:

"For example, trifluoroacetic acid (TFA) is found widely in rivers and has many sources, including the breakdown of pesticides, pharmaceuticals and refrigerants."⁶

Now the other document. The UK Pesticides National Action Plan, published in 2025, is the government's flagship statement on how pesticides are to be managed.

The word PFAS appears in it zero times.⁷

Two national plans. One government. The pesticides plan does not mention PFAS. The PFAS plan mentions pesticides once, and puts them in a river.

Neither document is a database. A database can be forgiven for speaking in chemical names. These two were written to be read.

The carbon that makes it a PFAS

Fludioxonil is licensed for use on fruit crops across Europe. Its name appears 160 times in that 2024 residue file.

The OECD set the test in 2021. It asks whether a molecule contains a carbon atom fully loaded with fluorine — no hydrogen, chlorine, bromine or iodine attached to it. Fludioxonil has one such carbon.⁸

That definition sorts molecules by shape, not by hazard. It was written that way deliberately, so that inventories and restrictions would have a boundary to work with.

Which means it gathers in thousands of substances that behave nothing like each other.

Fludioxonil is not PFOA.

It does not simply go away either. A 2026 review in PeerJ collects the published half-lives, and they run from days to years depending on where the substance sits.⁹

6 to 12 days on sunlit soil. 87 to 350 days in soil below the surface. About 120 days in sediments, and longer again in water. Once it is under the surface and out of the light, the review has its own word for it: "highly persistent."

Those are measurements of soil, water and sediment. They are not measurements of a human body.

On the fruit it is different again. It settles into the wax on the skin, where one storage study found the residue essentially unchanged after forty days.¹⁰

And breaking down is not the same as going away. That review makes a further argument. Fludioxonil breaks down into other substances, and those may be causing effects that have been blamed on the fungicide itself. It is a review advancing a mechanism, not new experimental work.⁹

The government's own plan says the same thing once. Its single pesticide sentence puts it generally: what turns up in rivers can come from the breakdown of pesticides, not just the pesticides themselves.

Fludioxonil is inside the definition anyway.

There is a second place it turns up in quantity. The Environmental Working Group, a US campaigning organisation, analyses the residue testing run by the US Department of Agriculture. Its 2026 review covered 54,344 samples across 47 fruits and vegetables.

In that review, fludioxonil was the pesticide detected most often — found in 14% of all the American produce tested.¹¹

That ranking is EWG's analysis of the government's samples. The USDA has published no ranking of its own.

American regulation has a version of the same gap. Fludioxonil is often applied to produce after it has been boxed for shipping — at that point it is working as a preservative. The FDA has not classified it as a food additive. So the Environmental Protection Agency regulates it as a pesticide and nothing else.⁹

The moulds it was made for

A fungicide on fruit sounds cosmetic. This one is not.

Fludioxonil is used against storage rots — the moulds that turn a stored apple to liquid over a winter. British apples and pears sit in cold stores for months, and some varieties are held for the better part of a year.

Nobody has published a number for how much fruit that saves. Take the most familiar of those moulds, Penicillium rot. NIAB studies apple storage disease in the UK. It says Penicillium rot is not necessarily responsible for large losses.¹²

What is established is the scale of the waste that already happens. WRAP reported in March 2025 that British households throw away more than 2.5 million tonnes of fresh fruit and vegetables a year. Of that, 1.7 million tonnes could have been eaten — about £4 billion of food.¹³

Less fungicide would mean more waste on top of that. Nobody has published a figure for how much, in either direction.

The benefit is real and it has never been added up.

EFSA says three things

The European Food Safety Authority publishes an annual account of residues across the food supply. In the most recent account, 98.8% of the 9,842 samples in the EU-coordinated programme complied with the law. Just over 43% carried no measurable residue at all.¹⁴

EFSA's own summary is that "the risk to human health from pesticide residues in food remains low."

That stands, and this report does not argue with it.

The second thing is a conclusion about the substance. In its 2024 peer review, EFSA identified fludioxonil as meeting the EU's criteria for an endocrine disruptor, for both people and wildlife. An endocrine disruptor is a substance that interferes with hormone systems.

Its own words describe the substance, not the amount on any piece of fruit: fludioxonil "meets the ED criteria for the oestrogen, androgen and steroidogenesis (EAS)-modalities for humans and non-target organisms."¹⁵

The third thing is what the same review says about its own completeness. EFSA records that "the consumer dietary risk assessment could not be finalised," citing gaps in the underlying data.¹⁵ There is more than one such gap in the document.

So the position is not a clean bill. Residues across the food supply are low-risk. This particular substance has been flagged as a hormone disruptor. And the sum of how much of it people actually eat was never finished.

The assessment that counts PFAS in food

One European instrument genuinely adds up PFAS in the diet.

In September 2020, EFSA added up the PFAS it had assessed and set a weekly limit for the total: 4.4 nanograms per kilogram of body weight.¹⁶ For a 70kg adult that is a few hundred nanograms a week — far less than a millionth of a gram.

It is a real threshold. Calculated, published, with a number attached.

The assessment also says where the exposure comes from. In its own words, these substances are "most often found in drinking water, fish, fruit, eggs, and egg products."¹⁶

Fruit is on the list. Third.

And the assessment covers four substances: PFOA, PFOS, PFNA and PFHxS.¹⁶ All four are legacy industrial chemicals — the long-lived, accumulating kind that built the word's reputation.

The word pesticide does not appear on that page. Neither does fludioxonil.

So a fungicide that meets the PFAS definition is licensed to be put on fruit. Its name appears 160 times in one year of the British monitoring file.

Nobody has estimated how much it contributes to a person's PFAS intake.

That is not an oversight somebody will get round to. The assessment that counts PFAS in food does not include the PFAS that is licensed to be put on the food.

Why the jar was clean

European law does contain a near-zero standard for pesticide residues. It exists and it is enforced.

It applies to processed cereal-based food and baby food for infants and young children. The limit is 0.01 milligrams per kilogram. That is effectively a rule against any detectable residue at all.¹⁷

That is the rule the jar of porridge was tested against. It is a detection threshold, not a health calculation — the level is set roughly where the instruments stop being able to see anything.

The adult limits are a different kind of number. EFSA's own exposure calculations for this substance use 4 milligrams per kilogram for strawberries, and 5 for apples and pears.¹⁸ Those are EFSA's working figures rather than the current legal limits — close, not exact.

Both are lawful. Neither implies that anyone broke a rule or that anyone was harmed. But they are not the same kind of limit, so dividing one into the other produces a number that means nothing: the infant figure marks where measurement begins, the adult figure marks where a health assessment landed.

They describe two regimes sitting one shelf apart in the same shop.

The jar and the fruit are in the same government file, under different rules, and one of them came back clean. One jar is one jar and proves nothing by itself. But most people have never been told there are two regimes at all.

Somebody has measured whether shoppers know. Once. One survey, in Germany, published in 2017. No follow-up appears to have been published since. That survey asked about pesticide residues in general, not about anything fluorinated.

In it, Germany's federal risk-assessment institute asked 1,004 people by telephone. 69% of organic shoppers and 61% of conventional ones believed pesticide residues in food are generally not permitted at all.¹⁹

Ten years on a list called Candidates for Substitution

Once you know the residue is there, the obvious question is why nothing has changed.

Fludioxonil and cyprodinil are both on a list called "Candidates for Substitution." The law says substances on it should be replaced where a better option exists.

Both went on the list when it was drawn up in March 2015. Both were still on it in the Commission's most recent consolidated version, dated July 2025. Ten years, no removal.²⁰

The list's name suggests a queue. The mechanism explains why it is not one.

To keep a substance on the market, nobody has to prove anything.

To restrict one, an authority has to demonstrate four things at once.

That a significantly safer alternative already exists. That substitution "does not present significant economic or practical disadvantages." That resistance in the target pest will not become a problem. And that the consequences for minor crop uses have been taken into account.²¹

All four. Not any one of them.

The law sets the economic condition high. To block substitution, the disadvantage has to be big enough to leave the grower unable to keep sufficient control of the pest — the law's phrase is "a major quantifiable impairment of working practices or business activity."

That condition is proved by showing something will not happen. And proving it falls to the regulator, not the manufacturer.

Which returns us to the committee. A draft regulation to move those approval dates has been before it since May.

Why not a PFAS column

The obvious demand at this point is a PFAS column on the government spreadsheet. That would be the wrong ask.

Fludioxonil is not PFOA, and the definition that catches it sorts by shape rather than by hazard.

A shopper reading "PFAS: yes" would think of the chemicals that built the word — the ones that accumulate and do not leave. Nothing known about this substance supports that comparison.

The reassuring read is not supported either. What fludioxonil does in a human body is not something the published assessments set out.

So the column would hand a reader certainty. The evidence supports none — not the alarming read, not the reassuring one.

Bolting a category onto a monitoring file does not make the category any easier to use. It moves the same unusable word somewhere new.

The gap is not a missing column. The translation already exists — a campaign group did it in 2024, from public data. Nothing inside government carries it across. Not into the plan for pesticides. Not into the plan for PFAS. Not onto the label the farmer reads. Not into the page that tells a shopper to go looking.

Parliament has asked. In April 2026 the Environmental Audit Committee reported that PFAS reach food through farming.²²

A farming network gave the committee the routes. Sewage sludge was the one it called most significant. Pesticides were on the list too, and it said they are often unlabelled.

The committee asked the government to set limits on PFAS entering the food chain that way.

The government replied in June. It wrote about strengthening testing for PFAS in fish, eggs, milk and vegetables, and about waiting for a toxicology review before considering maximum levels.²³

The word pesticide is not in the reply. Neither is residue, nor fruit, nor crop.

The same reply notes that a public website on chemical risks is "due to launch later this year." It had not launched.

One number would unsettle all this, and nobody has produced it. Someone could work out what fluorinated pesticide residues contribute to total PFAS intake. If the answer turns out to be a rounding error beside drinking water and packaging, this is a much smaller story — and we would say so.

Britain's own drinking water inspectorate already notes that water is "a small proportion of the possible overall exposure alongside use of various chemical products, food and even household dust."²⁴ The routes are spread out.

Nobody knows which route matters most for any given person. For fludioxonil, nobody has tried to find out.

Washing, peeling, organic

Washing does something, and less than the internet claims.

A one-minute rinse under the tap removes roughly a third of the fludioxonil from an apple. Soaking it in vinegar or baking soda gives no measurable advantage over plain water for this substance.²⁵

Worth doing. Not a solution.

Peeling reaches the residue. On fruit you have already bought, it is the only thing that does. Fludioxonil is not systemic — it does not travel into the flesh.²⁶ It sits in the wax on the skin, which is where the storage study found it still sitting after forty days.¹⁰

So the residue is in the peel, and peeling takes it off.

That has a price: you throw away part of the fruit you paid for.

You lose that for certain. You avoid an exposure the regulator calls low, and one nobody has measured for this route. There is no honest way to tell you whether that trade is worth making.

Strawberries cannot be peeled at all.

Organic avoids the question at source. Organic certification prohibits synthetic pesticide inputs, so the residue is not applied in the first place. That is a rule, not an inference from testing.

The direct sampling evidence is thin: PAN Europe's strawberry study included five organic and pesticide-free samples, and all five were clean. Five is five.²⁷

Organic also costs more, and that came up in the same parliamentary evidence. Dr Joanna Cloy of the environmental charity Fidra pointed to a "socio-economic divide" in who can afford organic produce.²²

The strawberry that took eight years

At Wageningen University in the Netherlands, researchers have spent eight years trying to grow strawberries without chemical crop protection.

They used a resilient cultivation system, natural pest control, and ultraviolet light against fungal disease instead of a spray. Kirsten Leiss, the plant health researcher who led the work, puts the result plainly: "We've shown that it's possible: a healthy strawberry grown with almost no chemicals."²⁸

Eight years, for one crop.

In 2025 the discounter Lidl began a two-year pilot in the Netherlands, cutting chemical crop protection across its supply chain. Its partners are the environmental organisation Natuur & Milieu and the agricultural consultancy CLM Research and Advisory. The crops are strawberries, potatoes and apples — and one of the techniques is that same ultraviolet light.²⁹

Neither of those things happened because a shopper asked for it. The research was done before anyone was buying, and the pilot began because a retailer wrote a different specification.

Whether the pilot holds past two years is not yet known. Neither it nor the Wageningen work is a policy instrument, in the UK or anywhere else.

The committee will decide something about fludioxonil before the end of September. It will extend the approval, or let it lapse, or do what it did in May and put the vote off again.

Whatever it decides will be written in the vocabulary of the residue file — a substance name, a regulation number, a date.

The blog post will still tell you to look for PFAS-free alternatives when you shop. The file will still name fludioxonil 160 times. And the only apple in it will still be a jar of baby porridge.