Spice food safety and contaminant limits

Spices are dried in the open, often on the ground, in the countries where they grow, and contamination with Salmonella from soil, animals, birds and handling can occur at any point before a treatment step. Salmonella does not grow in a dry spice, but it survives drying and storage for long periods, and it is markedly more heat-resistant in a low-moisture food than in a moist one. FDA sampling of imported spice shipments in 2007–2009 found Salmonella in 6.6% of them — about twice the rate across other imported foods — in shipments from 37 of the 79 countries sampled, so contamination is not confined to a few origins.

Spices are usually eaten in small amounts, but they are often added to foods that are not cooked afterwards — a salami coating, a dip, a snack seasoning — which is how spice contamination reaches people. FDA’s review identified fourteen outbreaks attributed to spices between 1973 and 2010, with 1,946 reported illnesses, 128 hospitalisations and two deaths, reported from nine countries.

Nothing a buyer can see or smell reveals Salmonella. The protection is a validated treatment in the supply chain, and cooking: spices added early in a cooked dish are heated with it, while spices sprinkled onto food that will not be cooked rely entirely on the treatment having been done.

Bacillus species live in soil and form spores that survive drying, storage and ordinary cooking, and they are common in dried spices. FDA’s review of spice-associated outbreaks includes a 2007 outbreak in France, with 146 cases, attributed to Bacillus cereus in the spice blend used in a couscous dish, and outbreaks linked to Bacillus subtilis and related species on turmeric-seasoned kebab in the United Kingdom and on peppered steak in New Zealand.

The illness comes from the bacteria multiplying, and in the case of B. cereus producing toxins, in food that has been cooked and then kept at warm temperatures. Spores are far more resistant to heat than Salmonella, so treatments designed to eliminate vegetative pathogens do not reliably remove them.

The kitchen control is time and temperature rather than the spice: cool cooked, spiced food quickly and do not hold it warm for hours, which is where these outbreaks come from.

FDA’s review of the literature from 1985 to 2012 lists Clostridium perfringens among the pathogens detected in spices, in more than twenty of them, among them black and white pepper, cinnamon, cumin, ginger, nutmeg, oregano and saffron. In 2024 two European rapid-alert notifications concerned C. perfringens in ground cinnamon, at up to 2.3 log CFU/g.

One outbreak, in France in 2007 with 19 cases, was attributed to herbs and spices, and FDA found another possible spice-associated outbreak that did not meet its inclusion criteria. The FAO’s 2022 risk assessment for spices and dried aromatic herbs names Salmonella and the spore-formers B. cereus and C. perfringens as the pathogens of particular concern, and FDA recommends more surveillance of C. perfringens in spices.

FDA’s review lists Staphylococcus aureus among the pathogens found in spices, reported in asafoetida, black and white pepper, capsicum, cardamom, cinnamon, garlic and ginger in the literature from 1985 to 2012. The forms of a spice were not always distinguished in those reports.

None of the fourteen spice-associated outbreaks FDA reviewed for 1973 to 2010 was caused by S. aureus, and a 2026 review of dried foods records no outbreak from herbs or spices. Its relevance is survival: in a study of inoculated dried spices and herbs held at room temperature in the dark, the time for a tenfold fall ranged from 6.1 days in allspice to 16.3 days in cinnamon.

FDA lists Cronobacter among the pathogens detected in spices, in anise seed and rosemary. A retail survey in Prague reported in a 2026 review found it in 4 of 10 basil, 3 of 10 ginger, 2 of 15 caraway, 2 of 12 marjoram and 2 of 10 allspice samples, and none in cinnamon, oregano or sweet pepper; a Warsaw survey found it in 9 of 38 samples of basil, parsley and tarragon. The review considers this unsurprising for an organism so widespread in the environment.

Cronobacter is considered to pose a risk mainly to infants under six months, especially newborns and premature infants, who are unlikely to eat these products. No outbreak has been linked to dried fruits, vegetables, herbs or spices, and no European rapid-alert notification for it in these products was found for 2020 to 2024.

European monitoring found Shiga toxin-producing E. coli in none of 119 herb and spice samples in 2011, and in 2 of 685 in 2019. Dried fenugreek leaves were the subject of two European rapid-alert notifications for it in 2020. Studies that looked for E. coli O157:H7 in selected spices, reviewed by FDA, did not find it.

Herbs and spices were implicated in a strong-evidence European outbreak of Shiga toxin-producing E. coli in 2012 and of another pathogenic E. coli in 2013. FDA names pathogenic E. coli as a target for more surveillance in spices because it can survive for long periods in low-moisture foods. The 2011 outbreak attributed to fenugreek seed concerned seed sold for sprouting, and the outbreak strain was not found in the seed.

Across the European monitoring studies a 2026 review gathers, 26 of 2,164 herb and spice samples, 1.2 per cent, carried Listeria monocytogenes, and every quantified result was below 100 CFU/g. A report of the organism in bay leaves, which FDA reviewed, was later clarified as concerning fresh rather than dried leaves.

No European rapid-alert notification or recall for L. monocytogenes in herbs and spices was found for 2020 to 2024, and none of the spice-associated outbreaks in either review was caused by it. Sample sizes and designs differed between the studies, so the prevalence figures cannot be compared directly.

FDA identifies drying as a point where mould can grow on spices that are not dried quickly or thoroughly enough, and storage as another, where spice that gets wet can grow Salmonella or mould. Mould filaments are among the filth elements FDA counts in spices.

FDA’s profile lists an import alert, in force since 1988, under which dried peppers from Mexico are detained without physical examination for excessive mould, and another, since 1987, covering Salmonella, filth, mould and foreign matter in Indian pepper. Shipments of whole dried ginger from China and of whole mild paprika peppers from China and Peru, refused for filth and mould, were accepted only after reconditioning by sorting or fumigation. The mycotoxins some moulds produce are recorded separately.

Aflatoxins are produced primarily by toxigenic strains of Aspergillus flavus and A. parasiticus, which grow on crops in warm climates and during slow drying and damp storage. The mould can be gone by the time the spice is sold; the toxin stays. Chilli, pepper, nutmeg, ginger and turmeric are the spices EU law names for aflatoxin limits.

Aflatoxin B1 is genotoxic and carcinogenic, with the liver as the critical organ, and EFSA considered it inappropriate to set a tolerable daily intake. In EFSA’s exposure assessment the main dietary sources are grains, grain products and milk rather than spices, which contribute little to total exposure because so little is eaten; the spice limits exist because contamination levels can be high.

No treatment that makes a spice safe from bacteria removes aflatoxin, and cooking does not destroy it. Control is upstream — fast drying, dry storage, and testing against the limits — and a buyer cannot detect it.

Ochratoxin A is produced by moulds of the genera Aspergillus and Penicillium — among them A. ochraceus, A. carbonarius and P. verrucosum — on crops that dry slowly or are stored damp. In EFSA’s occurrence data the highest mean concentrations were in chilli pepper, with pepper, nutmeg, ginger, turmeric and paprika among the other spice sources.

Its critical effect is on the kidney, and it causes kidney tumours in rats. Because the way it damages DNA is unresolved, EFSA withdrew its earlier tolerable weekly intake in 2020 and assesses ochratoxin A by margin of exposure instead.

Like aflatoxin, ochratoxin A survives the treatments used against bacteria and ordinary cooking. It is controlled by drying and storage at origin and by testing, and a buyer cannot detect it.

Lead in spices has two unrelated origins. Some lead reaches crops from the environment, and the EU sets maximum levels for dried spices category by category — seed, fruit, bark, root and rhizome, bud and flower-pistil spices. The serious cases are deliberate: lead chromate, a bright yellow industrial pigment, has been added to turmeric in Bangladesh to improve its colour, and in 2023 FDA traced lead and chromium in cinnamon applesauce pouches to cinnamon processed in Ecuador, where it concluded economically motivated adulteration was the likely cause. Samples of that cinnamon contained 2,270 and 5,110 parts per million of lead.

The WHO records no known safe blood lead concentration, and lead can permanently affect children’s brain development. The 2023 cinnamon incident produced 519 confirmed, probable and suspected cases in the United States, reported by CDC, from a single fruit-puree product.

Lead chromate cannot be seen, smelt or tasted at the levels used, and no home test separates it from turmeric’s own colour. Unusually vivid, cheap turmeric is a reason for caution rather than proof of anything; the protection is testing by the importer and the regulator.

Sudan I and related industrial dyes have been added to chilli products to make a dull or diluted powder look deep red. Their discovery in chilli powder used by food manufacturers in 2005 led to large withdrawals of products in Britain and across Europe and prompted EFSA’s review of the dyes illegally present in food.

EFSA concluded that Sudan I to IV, Para Red and Rhodamine B are or may be genotoxic and carcinogenic. None is permitted in food, so there is no acceptable level; any presence is a regulatory failure.

Sudan dyes are fat-soluble like chilli’s own pigments, so the popular test of stirring chilli into water cannot find them. They are detected by chromatography in a laboratory.

Ethylene oxide is a gas used to kill microbes and insects in spices and other dry goods. Residues remain in the treated food as ethylene oxide and as its reaction product 2-chloroethanol. The United States permits it as a postharvest fumigant on dried herbs and spices within set tolerances; the EU does not permit its use on food, and the Directive that authorised irradiation of spices in 1999 records that ethylene oxide could no longer be used to treat them.

FDA’s risk profile notes that toxic residues remain a concern with this treatment, and quotes a cancer-risk assessment that found the risk from residues at the concentrations found in spices "practically negligible". The EU’s approach is instead to limit residues to the level laboratories can detect, which is what made the 2020 discovery of ethylene oxide in sesame seeds from India a large incident.

Pesticides used on the growing crop, or in storage, can leave residues on the dried spice. The EU sets maximum residue levels for spices under its general pesticide-residue regulation, organised by spice group — seeds, fruits and berries, bark, roots and rhizomes, buds, flower stigmas and arils — and some, such as the level for ethylene oxide, are set at the limit of analytical detection.

FDA identifies pesticide residues among the hazards that good agricultural practice for spices is designed to prevent, alongside mycotoxins, heavy metals and undeclared colours. The hazard depends on the substance; residue limits are set so that compliant food is safe, and an exceedance is a regulatory matter rather than a measure of harm.

Polycyclic aromatic hydrocarbons form when wood and other organic material burn incompletely, and smoke deposits them on whatever it touches. Spices dried with direct contact with smoke or combustion gases take them up, and the EU limit for dried spices excludes smoked Capsicum — pimentón and chipotle, whose smoke is the point — and cardamom.

Several PAHs are genotoxic carcinogens. EFSA concluded that benzo[a]pyrene alone is not a suitable indicator and that a group of four PAHs is, which is why EU limits are set for benzo[a]pyrene and for the sum of four.

FDA found filth in 12% of imported spice shipments sampled in 2007–2009, 1.8 times the rate for other imported foods. The commonest adulterants were insect fragments, whole insects and animal hair, and nearly all the insects were stored-product pests, which points to poor packing and storage rather than the field.

Most filth is an indicator rather than a hazard in itself: rodent hair generally indicates contamination by droppings, and conditions that let pests in also let in pathogens. US law treats spice as adulterated when filth exceeds FDA’s defect action levels or when it was held in insanitary conditions.

Undeclared peanut and almond protein was found in ground cumin in 2014 and 2015, leading to recalls in North America and Europe and to reported allergic reactions; ground nut shells or flour used to bulk the spice is the explanation generally given. Blends, which combine many ingredients, carry the ordinary cross-contact risks of any mixed food.

For a person with a peanut or tree-nut allergy, an undeclared allergen is a direct hazard at very small amounts, and labelling law in the EU and the United States requires the named allergens to be declared.

Exposing spice to saturated steam for a controlled time so that its heat kills the microbes on it. Batch systems load packed spice into a chamber and inject steam, with or without pressure; continuous systems move unpacked spice through a steam chamber, some tumbling it so every particle is exposed. Vacuum-steam-vacuum designs draw air out so the steam penetrates, then draw the condensed water off again; others add indirect heat so the spice does not end wetter than it started.

To eliminate vegetative pathogens, above all Salmonella, before spice reaches food that may not be cooked. Properly designed and run, steam is an established and effective treatment, and FDA’s comparison of import and retail sampling is consistent with most contaminated shipments entering the United States being treated before sale.

Heat and moisture can drive off volatile aroma compounds, and FDA records that colour and flavour may be negatively affected by steam. How much depends on the system and the spice.

Colour can suffer with flavour, which matters most for the spices sold on their colour.

The moisture added in traditional steam treatment is the main quality cost, which is why newer designs remove condensed water under vacuum or control moisture with indirect heat.

Effective against vegetative bacteria such as Salmonella when every particle reaches the target temperature for long enough — which is not guaranteed, because Salmonella is far more heat-resistant in a dry food than in a moist one and inactivation can tail off. Bacterial spores are more resistant still. The treatment has to be validated for the spice and the system to be relied on.

• A system that does not expose every particle to steam for long enough, leaving surviving Salmonella
• Contamination after treatment, in handling, milling or packing
• "Steam washing", a cleaning step, sold or understood as a pathogen-reduction treatment
• Moisture left in the spice, which harms quality and storage

Exposing packed spice to ionising radiation — gamma rays from cobalt-60 or caesium-137, or machine-generated electron beams or X-rays — which kills microbes by damaging their DNA. The spice does not touch a radioactive source and is treated in its final packaging, so it cannot be recontaminated afterwards. US rules allow gamma sources, electron beams up to 10 MeV and X-rays up to 7.5 MeV.

Irradiation reduces microbial load with little heat or moisture, which is why it is used on spices whose aroma or colour steam would damage. FDA’s review describes it as cost-effective with minimal impact on the physical and chemical character of spices compared with steam or ethylene oxide, citing work that found no substantial change in the volatile oil of most spices at doses up to 15 kGy.

Little effect on volatile oil content at the doses reviewed, which is its main advantage over steam.

Irradiation does not make food radioactive, and FDA states that it does not noticeably change taste, texture or appearance at permitted uses.

An effective pathogen-reduction treatment for spices. The European Union permits dried aromatic herbs, spices and vegetable seasonings to be irradiated at a maximum overall average absorbed dose of 10 kGy — the only category on the EU list. The United States permits culinary herbs, seeds, spices, vegetable seasonings and their blends to be irradiated for microbial disinfection at up to 30 kGy.

• Consumer resistance, which FDA’s review names as the main disadvantage and which keeps it off many products where it would work
• Labelling that leaves buyers unsure whether an ingredient was irradiated

Treating packed spice in a sealed chamber with ethylene oxide gas, usually mixed with an inert gas, under controlled temperature and humidity, for several hours so that the gas penetrates the packs; the gas is then removed and the chamber flushed with air. Ethylene oxide kills microbes by reacting with their DNA and proteins.

It works without heat or moisture, which spares aroma, and it reaches into packed product. It is a long-established treatment for heat-sensitive materials, and the United States permits it on dried herbs and spices as a postharvest fumigant.

Ethylene oxide can cause chemical changes that affect quality, and it leaves residues of itself and of its reaction product 2-chloroethanol.

Its reduction of Salmonella in spices may be less than that of steam or irradiation, and it is harder to validate because more variables must be controlled — gas concentration, time, temperature, humidity, packaging and load. Dry organisms resist it more. The European Union does not permit it on food — the 1999 Directive listing spices for irradiation records that it could no longer be used to treat them — and limits its residues in spices to the level of quantification.

• Packaging the gas cannot penetrate, such as foil-lined film, making the treatment ineffective
• Residues of ethylene oxide and 2-chloroethanol above the limits of the destination market
• Treatments that were never validated for the spice and load

Treatments studied as alternatives to steam, irradiation and ethylene oxide: dry heat, microwave heating, high pressure, supercritical carbon dioxide, ozone, pulsed light, and a steam variant called controlled condensation.

Each tries to reduce microbes with less damage to aroma and colour or without the regulatory and consumer objections to the established treatments. FDA found them not in use, or only minimally in use, on spices commercially.

Radio-frequency heating is the most studied of them on spices. In laboratory trials, black peppercorns heated for two and a half minutes showed no significant change in colour, piperine or most of their volatile compounds, and the colour of black and red pepper heated for under a minute was unaffected. These are results on conditioned laboratory samples, read in summary.

Their effectiveness on spices is less established than that of the three main treatments, and any of them would need validating for the spice and the equipment before it could be relied on.