Start with why soft fruit gets sprayed at all

There is a reason the residue rankings are full of berries and leafy greens rather than avocados and pineapples. It is not a conspiracy. It is agronomy. Soft fruit is close to the hardest thing in the produce aisle to bring to market intact, and three features of it drive the whole spray programme.

• There is nothing to peel. A strawberry is edible surface the whole way round. This is the mechanism the Environmental Working Group itself points to: with no protective skin, residue can absorb into the flesh rather than sitting somewhere you can remove.⁵
• It rots on a clock. Strawberries have a marketable window of roughly seven to ten days even when they are handled well, and grey mould is the defining loss in storage.¹⁵ Most of the fungicide load on soft fruit exists because of that clock.
• It is picked again and again. Berries do not ripen all at once, so a field is harvested repeatedly across weeks. Spraying and picking overlap in a way they never do for a grain crop.

Every one of those follows from one demand: perfect looking soft fruit, year round, at volume, thousands of kilometres from where it grew. The residues are downstream of that. So is everything else in this article.

What the testing actually found

The Environmental Working Group publishes an annual Shopper's Guide built on the United States Department of Agriculture's Pesticide Data Program. The 2026 edition covered 54,344 samples across 47 fruits and vegetables, ranked by how often residues appeared, how many different ones appeared together, how concentrated they were, and how toxic each compound is.¹²

Here is the list itself. The column on the right is this article's addition, not the Environmental Working Group's, because it turns out to explain most of the ranking on its own.

Two things in that table are worth slowing down for. The first is that strawberries are no longer number one, and the reason is interesting. They held the top spot for nine straight years until 2025, when the ranking was rebuilt to weight the toxicity of each individual pesticide rather than only counting how many were present and at what concentration.⁴ Spinach took the lead under the new maths and has kept it.

The second is that being lower on the list is not a clean bill. Blueberries sit at rank twelve, which sounds reassuring until you remember they still made a list of the twelve worst out of forty seven. And the specific strawberry data is worth seeing on its own, because it is the clearest picture of what "sprayed" means in practice.

The criticism, printed next to the data

This ranking is contested by working toxicologists, and any article that quotes it without saying so is selling you something. So here is the strongest version of the objection, from the paper that made it best.

In 2011, Carl Winter and Josh Katz at the University of California, Davis ran a probabilistic exposure assessment on the ten most frequently detected pesticides on each of the twelve listed commodities, using the same USDA data the list is built from. All 120 exposure estimates came in below the established chronic reference doses. Only one of the 120 exceeded 1 percent of the reference dose, and that was methamidophos on bell peppers, at 2 percent.⁶ Their conclusion was that substituting organic for conventional produce in these twelve items did not produce a meaningful reduction in consumer risk.

The methodological point underneath that is simple and fair. The Dirty Dozen counts how often residue is detected and how toxic the compound is in principle. It does not run a dose response assessment on the person eating the fruit. Detection is not exposure, and exposure is not harm. Analytical chemistry has become sensitive enough to find almost anything, and finding it does not mean it matters.

Residue data tells you what is on the fruit. It does not tell you the fruit is dangerous. Both halves of that sentence matter.

So this article will not tell you strawberries are poison. What the list is genuinely good at is the narrow thing it measures: where residue concentrates, and where it stacks. If two foods offer comparable nutrition and one of them consistently arrives carrying ten compounds instead of one, that is a real input into a shopping decision. It is one input, and it is not the biggest one in this article.

Why washing only goes so far

This is where almost every article on the subject fails, in one direction or the other. Either washing is sold as a complete answer, or it is dismissed as pointless. The evidence supports neither, and the real answer is more useful than both.

A 2026 scoping review in Frontiers in Environmental Health pulled together 47 peer reviewed studies covering 23 produce types and 79 pesticides or metabolites, about 1,100 data points, and reported a median reduction for each household method.⁷

Read those ranges again. Zero to ninety nine percent, for the same method, on different produce and different compounds. That spread is not noise. It is the single most useful piece of knowledge in this article, and it comes down to one distinction.

A contact pesticide sits on the outside of the fruit. Surface treatment reaches it, which is where the ninety percent results come from. A systemic pesticide is taken up by the plant and travels through its tissue. The review names azoxystrobin, tetraconazole and acetamiprid, and states plainly that a portion of a systemic residue can become unreachable by washing.⁷ It is not on the fruit. It is in the fruit. There is no soak, scrub, peel or spray bottle that gets it out, because it is part of the tissue you are eating.

Which makes the honest instruction narrower and more actionable than the usual advice. Soaking beats rinsing by roughly twenty percentage points, it costs nothing but ten minutes of leaving something alone, so do it. And understand what you have done: you have addressed the surface fraction. The rest was never on the surface.

The fruit was bred for the truck

Residue is the visible half of the story. The invisible half is that the fruit itself was selected, over decades, for traits that have nothing to do with what is inside it. Yield. Size. Firmness. Uniform colour. Days on a shelf. Those are what the supply chain pays for, so those are what got selected.

The clearest proof of what that costs is in tomatoes, not berries, and it is one of the better pieces of plant science of the last fifteen years. Nearly every tomato grown for the American grocery trade carries the uniform ripening mutation, selected because it makes fruit ripen to an even colour, which is what makes machine timed harvest and identical looking retail fruit possible. In 2012, a team publishing in Science showed that this mutation works by disabling a transcription factor called GLK2, which drives chloroplast development in the developing fruit. Losing it reduced the sugars and the carotenoids in the ripe tomato.⁸⁹

Read that again slowly. The trait that made supermarket tomatoes look uniform is the same trait that took out part of their sweetness and part of their pigment chemistry. Nobody chose that. Nobody knew. They selected for the thing they could see, and something they could not see came along with it.

For the wider crop picture there is one much quoted analysis, and it is worth stating accurately rather than the way it usually travels. In 2004, Davis, Epp and Riordan compared USDA composition data for 43 garden crops between 1950 and 1999. They found statistically reliable declines in six of thirteen nutrients: protein, calcium, phosphorus, iron, riboflavin and ascorbic acid. Median declines ran from 6 percent for protein to 38 percent for riboflavin.¹²

You will see this study quoted as proof that food has lost most of its nutrition. It is not that, and pretending otherwise would undercut everything else on this page. Seven of the thirteen nutrients showed no reliable change at all, and a meaningful share of the individual changes were increases. The defensible version is narrow and still worth knowing: the varieties changed, they changed toward yield and durability, and where composition moved, it mostly moved down.

What happens after it is picked

The last part of the system is the part nobody photographs. Commodity fruit is not picked when it is ready. It is picked when it can survive the trip, and finished later on a schedule.

Bananas are the clearest example because the process is completely documented. They are harvested green on purpose, then held up to about twenty days in refrigerated shipping containers at roughly 13 degrees Celsius, a temperature chosen specifically to stop them ripening in transit. On arrival they go into sealed, gas tight ripening rooms, typically thirty to forty feet long. There they are exposed to 100 to 150 parts per million of ethylene for 24 to 48 hours at 15 to 20 degrees, at 90 to 95 percent humidity, from a commercial mix of about 4 percent ethylene in nitrogen. Then they colour up over roughly four to eight days, timed to whatever ripeness stage the retailer ordered.¹³¹⁴

Now the part that most articles about this get wrong on purpose. Ethylene is the plant's own ripening hormone. The fruit makes it itself. Commercial ripening applies the same molecule the banana would have produced on the tree, and organic bananas go through the same rooms. This is not a toxicity story and anyone selling it to you as one is not being straight.

The real point is structural, and it is worth more than the scare version. Ripeness in a supermarket is a logistics output. It is a date the supply chain hit, not a signal that the fruit finished developing on the plant. Berries cannot even be handled that way, because they do not continue ripening after picking, which is why the marketable window is seven to ten days and why they are picked as late as the journey allows and not a day later.¹⁵

And the clock keeps running after that. Storing fruits and vegetables at 4 degrees for two weeks significantly decreased their vitamin C, with tomato losing the most at 71.8 percent, cut total phenolics by up to 29.67 percent, and dropped antioxidant capacity along with them.¹⁶ Strawberries specifically lose vitamin C, firmness, phenolic compounds and anthocyanins throughout storage.¹⁵ By the time a punnet reaches a shelf, days of that have already happened.

You will also see a number circulating that spinach loses 29 percent of its vitamin C in a single day and 94 percent in a week. We could not trace it to a published paper, so it is not in this article. The direction is well established. Those particular figures are not.

Which leads somewhere useful and slightly counterintuitive. A 2015 study compared fresh stored and frozen produce across eight fruits and vegetables, and found ascorbic acid showed no significant difference in five of the eight, and was actually higher in the frozen samples in the other three.¹⁷ Frozen fruit is picked riper, because it does not have to survive a journey looking good. For anything that travelled a long way to reach you, frozen is very often the better buy, not the compromise.

The foods that sidestep the whole system

Everything above is a description of one production model: high volume, long distance, appearance graded, shelf life optimised. There is a category of food that simply is not in that model, and it is the most practical thing you can take from this article.

Wild harvested and low volume berries do not appear in residue rankings, largely because they are not intensively sprayed commercial crops. Lingonberry is overwhelmingly wild picked across the boreal forest and Arctic tundra.¹⁸ Cloudberry is almost entirely wild harvested, because commercial cultivation of it has mostly failed.¹⁹ Sea buckthorn grows wild on the Tibetan Plateau above 3,000 metres, about 10,000 feet, and in Inner Mongolia, on a thorned shrub that fixes its own nitrogen and survives where very little else does.²¹

Two honest caveats, because the point of this article is to be right rather than convenient.

First, absence from the Dirty Dozen is not evidence of purity. These berries were not tested and cleared. They were not among the 47 produce types the USDA programme covers, because almost nobody in North America eats them at volume. Not on the list means not measured, and those are different claims.

Second, the moment a wild berry becomes a powder or a capsule, you are back to trusting a label. That is not a small risk. In an authenticity study using nuclear magnetic resonance profiling, only 3 of 10 commercial acerola supplements had a metabolomic profile consistent with genuine acerola powder, and several showed ascorbic acid to choline ratios more than ten times what the real fruit gives, which is the signature of synthetic vitamin C being added to the pot.²⁰ Seven in ten did not look like the fruit on the front of the packet.

Which is the whole thing in one line. Turn the packet around. If the ingredient list says anything other than the fruit, you are not buying the fruit.

The bottom line

The berries in the middle of the supermarket are not poison, and the residue ranking that says they carry the most does not claim they are. What that ranking is really measuring, without meaning to, is a production model: fruit with no skin, grown at volume, bred to survive a truck, picked before it is finished, ripened or chilled on a schedule, and priced on how it looks when it lands.

You cannot wash that off, because most of it was never on the surface. What you can do is soak what you buy, spend the organic premium where the skin is the food, buy frozen when the food travelled, and widen what counts as a berry beyond the two that got famous.

Frequently asked questions

Keep reading

The 10 best berries in the world, ranked

Berries you have never heard of

Sea buckthorn, the king of berries