Are Pistachios Poisonous? What the Law Actually Limits

No. A pistachio is the seed of Pistacia vera, and it carries no toxin of its own. Neither does the shell, which is woody and indigestible rather than dangerous. The only hazard food regulators put a legal number on is a mold toxin that can grow on the crop, and not anything the plant itself makes.

That number turns out to be the interesting part. We compiled it across the four bodies that set one, and they do not agree. The loosest limit is two and a half times the strictest, on the same nut.

The legal aflatoxin limit for pistachios, by jurisdiction

Compiled 11 September 2026 from the primary legal texts, linked in full under sources. Figures are micrograms per kilogram, which is the same as parts per billion. Downloadable as a CSV file, free to reuse with attribution (see citing this page).

Read across the rows and four things stand out.

The United States regulates this with one number, and Europe with four. FDA's guidance for pistachios is a single sentence: it may consider pistachio nuts containing aflatoxins greater than 20 micrograms per kilogram to be adulterated. That is total aflatoxins, with no separate figure for aflatoxin B1, the most potent of the four, and no distinction between a batch heading straight to a shelf and one that will still be picked over. The EU sets a limit for B1 on its own, a second for the total, and splits both according to whether the nuts will be sorted first.

The strictest figure in the table is 8.0 and the loosest is 20. A consignment carrying 9 micrograms per kilogram of aflatoxin B1 is legal to sell in an American supermarket and illegal in a Spanish one. Nothing about the nut changes across that line.

The US sets no limit at all for ochratoxin A in pistachios. The EU sets two, and it names pistachios alone in both of them. That is worth pausing on, because in the aflatoxin entries immediately above, the same regulation lumps pistachios in with almonds and apricot kernels. For ochratoxin A, pistachio gets its own line in European law and nothing in American law.

The looser limit for unsorted batches is the regulation admitting that sorting works. Aflatoxin does not spread evenly through a lot. It concentrates in a small number of damaged, moldy or insect-bored nuts, which is why removing those nuts pulls the average down so sharply, and why both the EU and Codex write two numbers instead of one.

What aflatoxin actually is, and how worried to be

Aflatoxins are made by molds in the Aspergillus genus, mainly A. flavus and A. parasiticus, which grow on crops in hot weather and in damp storage. The World Health Organization's list of foods they affect runs through cereals, oilseeds and spices before reaching tree nuts, where it names "pistachio, almond, walnut, coconut and Brazil nut", in that order. Pistachio comes first on that list, and the reason is climate: the crop grows in exactly the hot, dry places where a wet spell at the wrong moment does damage. Our pistachio prices tracker covers what those weather years do to the supply.

The health picture has two halves and they are very different sizes. Large doses cause acute aflatoxicosis, which WHO describes as damaging the liver and potentially life threatening. That is a real phenomenon in places where a staple crop fails badly and contaminated grain is eaten in quantity. It is not what a bag of pistachios from a supermarket does. The half that the limits in the table are aimed at is the chronic one: aflatoxins are genotoxic, and WHO states there is evidence they cause liver cancer in humans. Regulators set low numbers because the risk accumulates, not because any single serving is dangerous.

Ochratoxin A, the second toxin in the table, is produced by Aspergillus and Penicillium molds. WHO calls kidney damage its most sensitive and notable effect.

Does roasting help?

Possibly, and pistachio is the nut with the best evidence for it. Codex says so in an unusually direct aside while defining which nuts count as ready-to-eat: shelling, blanching and sorting are the treatments proven to reduce aflatoxin, and then it adds that "there is some evidence that roasting reduces aflatoxins in pistachios but for other nuts the evidence is still to be supplied."

That is a narrow claim and worth keeping narrow. It does not mean roasting makes a contaminated batch safe, and no regulator treats roasting as a control step the way it treats sorting. If you want the rest of what heat does to the nut, we cover it in roasted vs raw pistachios and how to roast pistachios.

Are raw pistachios poisonous?

No, and the confusion is borrowed from a different nut. Cashews are never sold in their shells, because the cashew shell carries a caustic oil, so a cashew labeled raw has already been steamed or roasted to drive that oil off. Pistachios need no equivalent step. They are sold, shipped and eaten in the shell, and Codex's sampling plan takes this as read: in the same procedure that specifies shelled laboratory samples for almonds, hazelnuts and Brazil nuts, the pistachio sample is specified in-shell, because that is how the nut is predominantly traded.

The family connection behind the rumor is real, though. Pistachio, cashew, mango, poison ivy and sumac all sit in the Anacardiaceae. That shared ancestry shows up in the kitchen as nothing at all, and in allergy clinics as a great deal, which is the next section.

Pistachio allergy is the real answer to most of these searches

Pistachio is a tree nut for labeling purposes on both sides of the Atlantic, and an allergic reaction is what most people are describing when they reach for the word poisoning. The cashew relationship is the strong one in the literature. In the IDEAL study of cashew-sensitized children, 98 percent were also sensitized to pistachio, and mango co-sensitization ran at 21 percent.

The figure that matters more sits next to it. Pistachio sensitization was clinically relevant in only 34 percent of those children. Sensitization means a blood or skin test picked up antibodies. Clinical allergy means the food actually causes a reaction, and the two come apart often enough that the study's authors recommend a supervised food challenge rather than blanket avoidance. Roughly one in five children allergic to one of the two nuts tolerates the other.

None of that is something to work out at home. If you suspect a nut allergy, that is a conversation with a doctor, and a supervised challenge is the test that settles it.

The other things people mean by a bad pistachio

Rancidity. Pistachios are about half fat, and fat goes off. A rancid nut tastes sharp, bitter and a little like old paint, and while it is unpleasant rather than poisonous, there is no reason to finish it. Cool, dark and airtight storage buys months.

Visible mold. Discard it, and discard the ones touching it. This is the one spoilage sign that connects directly to the table above.

Shells. Not toxic, but genuinely not food. They are woody, they do not break down, and eaten in quantity they can obstruct the gut. If you are getting through a lot, how to shell pistachios will speed things up.

Red dye. Bright red pistachios are a historical habit rather than a hazard: the color was used to hide stains picked up during harvest, back when harvesting was done by hand. Red pistachios has the full story.

Salt. A practical issue rather than a safety one, but a real one if you are watching sodium, and we compare the options in salted vs unsalted pistachios.

For what pistachios actually do for you, rather than what they might do to you, the evidence review sits on are pistachios good for you, which registers every published human feeding trial and what each one was funded by.

How we compiled this, and what it leaves out

• Every figure comes from the primary legal text, not from a summary, a news article or a secondary database. The FDA figures were read from the two Compliance Policy Guide PDFs; the EU figures from the consolidated text of Regulation (EU) 2023/915 on EUR-Lex; the UK figures from the retained text of Regulation 1881/2006 on legislation.gov.uk; and the Codex figures from CXS 193-1995 as published by FAO. All were checked on 11 September 2026.
• The EU publishes these figures with a comma as the decimal mark ("8,0"). We have written them with a point to match the rest of the table and the American audience. No value has been converted or rounded.
• The UK ochratoxin A cells say "not compiled" rather than "no limit". We read the aflatoxin entries in the retained Annex and did not verify the ochratoxin A entries for pistachios, so we have left the cells empty rather than assume the UK matches the EU. Absence of a figure here means we did not check, not that no limit exists.
• The UK text is the retained version published on legislation.gov.uk, which tracks amendments from the end of the implementation period on 31 December 2020. We have not independently confirmed whether Great Britain has amended these particular entries since, so treat the UK rows as the retained baseline.
• FDA action levels are guidance, not regulation. The Compliance Policy Guide says so on its face: it contains nonbinding recommendations and does not have the force of law. It describes when FDA may consider a food adulterated under the statute. EU and UK maximum levels are binding law. Putting them in one table compares what each authority enforces to, which is the useful comparison, but the two are different kinds of instrument and the table should not be read as saying otherwise.
• We did not compile enforcement data. How often pistachios are actually detained, refused or recalled for aflatoxin is a separate question from what the limit is, and it needs a different dataset.
• Jurisdictions outside these four are not covered. China, Japan, India and the Gulf states all import pistachios in volume and all set their own limits. This table is not a world survey.

Spot an error, or have a primary citation for a cell we left empty? Email us via the about page and we'll correct it and note the change here.

Citing this page

The compiled table and dataset are original work. You're welcome to reuse them, including for commercial publication, with attribution to pistachiomilk.org and a link to this page (Creative Commons CC BY 4.0). A suggested citation:

Ricci, E. (2026). Are Pistachios Poisonous? What the Law Actually Limits. pistachiomilk.org. https://pistachiomilk.org/are-pistachios-poisonous/

Version history: v1.0, 11 September 2026. First publication of the compiled aflatoxin and ochratoxin A limits table (17 dataset rows across four jurisdictions). Corrections and additions will be listed here.