Guide · Produce
Washing produce is two completely different jobs, and almost every guide online mashes them into one. Dirt and germs come off in the sink. Pesticide residue is engineered not to — that's the entire point of the formulation.
Everything on this page is sorted by which job it's doing. If a method only helps with one of them, it says so.
This part is settled, and it's worth doing on every single thing you buy.
This is the part that gets skipped, and it's the part that actually explains everything.
A pesticide that washed off in the rain would be useless. So the industry solved that, on purpose, and there's a word for it: rainfastness. Formulations carry sticker adjuvants — polymers whose chains bond the pesticide particle to the leaf surface and hold it there through rainfall. The trade literature is completely open about the design goal, and there are two routes to it:
Then there's the category where the question doesn't even apply. Systemic pesticides — neonicotinoids are the ones most people have heard of — are taken up by the roots or the seed and moved through the entire plant. The residue is in the tissue. As one review puts it plainly, they cannot simply be washed off or peeled away before eating. There's no sink technique for that. There's no technique at all.
And the field data agrees. The most realistic study anyone has run on this is from the Connecticut Agricultural Experiment Station — pesticides applied under normal field conditions, left to weather naturally on the crop before harvest, then rinsed under tap water. Rinsing did reduce 9 of the 12. But vinclozolin, bifenthrin and chlorpyrifos came off at zero — no measurable reduction at all.
The finding underneath that is the useful one: how well a pesticide rinses off has no relationship to how water-soluble it is. It's about how tightly the compound has bound into the waxy cuticle of the skin. Once it's sorbed into that wax layer, water is the wrong tool no matter how long you run it.
You'll see big removal percentages quoted everywhere. Here's what's behind them, because the methods matter more than the headline.
The number I'd treat with the most suspicion is the one you see quoted most. A 2022 comparison is the source of the popular "running water removes 77% of pesticides, better than baking soda or vinegar" claim. Its method: leafy vegetables dipped in a mixed pesticide solution for 10 seconds, then air-dried for 15 hours.
That's pesticide in solution, deposited onto a picked leaf a few hours earlier. No commercial formulation, no sticker adjuvant, no pre-harvest interval, nothing sprayed on a living plant and weathered for weeks. It's a clean experiment for ranking methods against each other, and a bad basis for telling you what comes off a supermarket vegetable. I had that number sitting at the top of this page. It shouldn't have been.
The apple study behind the baking soda advice — mine included — is the more useful one, and it's also the more honest about its own limits. UMass Amherst applied a fungicide and an insecticide to organic Gala apples, then compared tap water, a 1% baking soda solution, and an EPA-approved bleach solution.
Baking soda beat both tap water and bleach, and that finding holds. But read the third number. The fungicide had already migrated 80 micrometres into the flesh, and the researchers were direct that no washing method removed the internalised residue — only peeling did. Surface removal was never the whole problem.
So where does that leave you? Three honest conclusions:
1. Wash everything, every time — that job is real and water does it. 2. For residue on firm, waxy produce, the soda soak is the best sink-level option anyone has measured, and it's still only working on what stayed on the outside. 3. When residue is the actual worry, peeling and buying differently do more than any wash. That's not a satisfying answer. It's the one the data supports.
Every entry answers both jobs: Rinse is the dirt-and-germs step and you do it on everything. Soda soak is the residue step — the badge tells you how long that item can take, because the limit is what the produce can physically tolerate. Three things genuinely shouldn't soak, and they say so.
The label on each one tells you what it's for. A method that's excellent against bacteria can be near-useless against residue, and that distinction is the whole reason this page exists.
If you're going to do it, do it at the concentration and the duration that was actually tested. Most versions online get the time wrong.
That's roughly a 1% solution, which is the concentration that was tested. The rinse afterwards isn't optional — you're washing off the baking soda along with what it lifted.
Use baking soda, not baking powder. The mechanism is alkalinity — the solution breaks the pesticide molecule down. Baking powder is baking soda already buffered with an acid, plus cornstarch. Same shelf, different chemistry, and it won't do this job.
How long, by what it is. The 12–15 minutes everyone quotes is the apple protocol. Softer produce can't take that, so it gets less — the soak still does something, you just stop before the food suffers.
12–15 min — apples, pears, cucumbers, bell peppers, hard squash, citrus you're
going to zest.
8–12 min — grapes, firm stone fruit.
3–8 min — berries, tomatoes, leafy greens, broccoli, celery, asparagus, herbs,
potatoes you're eating the skins of.
Don't soak — raspberries (hollow, they fill), mushrooms (a sponge, they drink it),
and anything pre-washed or pre-cut.
Drying matters more the shorter the soak. On berries especially, a soak you don't dry off properly costs you three days of shelf life.
What this soak is and isn't. It is the best-performing surface treatment anyone has measured at kitchen scale. It is not a decontamination step. It works on what stayed on the outside of the skin — not on what bound into the wax, not on what migrated into the flesh, and not at all on systemics. Do it because it's the best available option for firm produce, not because it makes the fruit clean.
Two of these are actively worse than doing nothing.
And some perspective, because I'm not here to make you scared of vegetables. The USDA tests thousands of samples a year for exactly this. In the 2024 Pesticide Data Program summary — 9,872 samples — over 99% came in below the EPA's limits, and 42.3% had no detectable residue at all. Residues over the legal tolerance turned up in 0.77% of samples, and 63 of those 76 were imported.
Wash your produce. Then eat a lot of it. The risk of the vegetable you don't eat is much better documented than the risk of the residue on the one you do.