Start with the size of a serving

Federal regulation sets a reference serving for spices and herbs, and it is worth knowing before anything else:

"Spices, herbs (other than dietary supplements)": ¼ teaspoon, or 0.5 g if not measurable by teaspoon.

Half a gram. A whole teaspoon is roughly two grams.

That number settles a lot of arguments. At two grams, a teaspoon of any dried herb or spice contributes a fraction of a gram of fiber, a rounding error of protein, and trace amounts of minerals. Spices are not a nutrition source, and no amount of enthusiasm changes the arithmetic.

This is not a criticism. Nobody eats oregano for the fiber. It is a reason to be suspicious of any article that presents a nutrient table for spices as though a teaspoon were a portion: the table is technically accurate per 100 grams and meaningless at the amounts anyone uses.

The antioxidant numbers were withdrawn

For about a decade, spice marketing ran on a single measurement: ORAC, or Oxygen Radical Absorbance Capacity. Cloves scored astronomically. Cinnamon scored high. Charts circulated ranking foods by ORAC value, and those charts are still circulating.

The USDA database those numbers came from is no longer published. Every USDA Agricultural Research Service address that once hosted it now returns an error, and secondary sources report that the agency withdrew it in 2012. The withdrawal statement itself could not be retrieved for this article, so neither its wording nor the agency's stated reasoning is given here, only that the values are no longer published by the body that produced them.

What can be cited is the peer-reviewed position on the whole category of measurement. A 2014 paper in Nutrition concluded that "the use of TAC assays for the in vitro assessment of antioxidant quality of food, which often is employed as a marketing argument or for the assessment of the 'wholesomeness' of food, is to be discouraged."

A 2020 review in the Journal of the Science of Food and Agriculture was blunter: "A direct link between a food TAC value and any health benefit is erroneous and has led to several cases of consumer deception."

The problem is not that the measurements were wrong. It is that they measure how a substance behaves in a test tube against a chemical reagent, which tells you nothing about what happens after you eat it, absorb a small and variable fraction of it, and metabolize the rest.

So: ignore ORAC numbers entirely, for any food, for any purpose. If a product is marketed on one, that tells you something about the marketing.

What can be said about spices and health

Very little, and this article is not going to pretend otherwise.

The cinnamon research is almost entirely in people with type 2 diabetes, a diagnosed medical condition, and it is a mess. A 2025 review reported heterogeneity statistics of 88 percent for fasting glucose, 100 percent for post-meal glucose and 94 percent for HbA1c, which is another way of saying the studies disagree so completely that pooling them produces a number without a meaning. A 2024 analysis found the insulin result not statistically significant (p = 0.058), and its authors did not claim a demonstrated effect. And a 2019 analysis found HbA1c, weight, BMI, waist circumference, insulin and insulin resistance all null, with its authors writing that "due to high heterogeneity findings should be interpreted with great caution."

There is also an independence problem: several of these analyses come from a small set of research groups with overlapping authorship, which means agreement between them is not the same as replication.

Turmeric has a more specific problem. A 2017 paper in the Journal of Medicinal Chemistry, examining curcumin as a drug candidate, classified it as both a PAINS compound (pan-assay interference: a molecule that produces false positives across unrelated laboratory assays) and an "invalid metabolic panacea." Its blunt summary: "No double-blinded, placebo controlled clinical trial of curcumin has been successful." The authors describe it as "an unstable, reactive, nonbioavailable compound and, therefore, a highly improbable lead."

In fairness, that paper is about curcumin as a pharmaceutical, not about turmeric in a curry. Nobody is claiming the spice in your food is harmful. But it is the most rigorous look at the molecule's behavior, and it should temper anyone's expectations about what a spice can do.

This article therefore makes no health claims about any spice, because the evidence does not support them, and a claim that outruns its evidence is worth less than no claim at all.

The one thing that is supported

There is a real, useful finding, and it is about salt.

A randomized trial published in 2015 tested a behavioral program designed to help people reduce sodium, in which herbs and spices featured prominently as replacements. Twenty-four-hour urinary sodium (a proper measurement, not a food diary) was lower in the intervention group by about 957 milligrams a day (95% CI 375 to 1,539; P=0.002).

Read the limits carefully. Only 40 people were randomized. It was a single site, 88 percent African American, mean age 61, with most participants hypertensive. And it was a multifactorial behavioral program (cooking instruction, group support, self-monitoring) in which spices were one component among several. You cannot pull the spices out of it and attribute the result to them.

The defensible statement is that a behavioral program in which herbs and spices featured prominently helped participants sustain a lower sodium intake. Not "spices lower your sodium," and nothing at all about blood pressure.

Even hedged that heavily, it is the most substantive thing anyone can say about herbs and spices, and it points at the right use: they are what makes low-salt food taste like food.

The safety issues nobody mentions

Two, and they are more interesting than the health claims.

Lead in turmeric

This one is real, documented, and specific.

A 2024 study in Science of the Total Environment tested 356 turmeric samples from 23 cities across India, Pakistan, Sri Lanka and Nepal. Fourteen percent had lead above 2 micrograms per gram. In the worst samples (at or above 18 µg/g) the ratio of lead to chromium was close to 1:1, which is the chemical signature of lead chromate, a bright yellow pigment sometimes added to turmeric to improve its color. Some samples exceeded 1,000 µg/g.

This is adulteration, not contamination: someone put it there.

It is also demonstrably fixable. A companion 2023 study found that an enforcement campaign in Bangladesh dropped the proportion of market turmeric with detectable lead from 47 percent to zero (n=631, p<0.0001). That is one of the more encouraging public-health results in this whole subject.

What it means for a shopper: the risk is concentrated in specific supply chains rather than distributed across your spice rack, and it is the reason sourcing and testing matter for turmeric more than for most spices. It is not a reason to treat spices generally as hazardous.

One thing that does not apply here is FDA's Closer to Zero program, which is sometimes invoked in this context. It targets foods commonly eaten by babies and young children, and it does not cover spices generally.

Salmonella, and the reassuring comparison

FDA has studied pathogens in spices in detail, in a risk profile first issued as a draft in 2013 and updated in 2017.

At import, the numbers are attention-getting. Across fiscal years 2007 to 2009, the average shipment prevalence of Salmonella in imported spice was 6.6 percent, and the prevalence of filth adulteration was 12 percent. By spice type in later years, prevalence ranged from 0 to 18 percent.

At retail, the numbers are completely different. A 2017 study testing 7,250 retail samples found prevalence of 0.00 percent for cumin, sesame and white pepper, and 0.15 to 0.64 percent for the other eight spices tested.

That gap, up to 18 percent at the border against well under 1 percent on the shelf, is the most useful fact in this section. The retail study made the comparison directly, putting its own figures against 1.7 to 18 percent at import for the same spices. The frightening import number describes shipments being screened at entry, many of which are then treated or rejected. It does not describe the jar in your cupboard.

Two findings cut the other way and belong here for balance. First, FDA's profile found that "Consumption of ready-to-eat foods prepared with spices applied after the final food manufacturing pathogen reduction step accounted for 70% of the illnesses." Spices added to something already cooked are the pathway that matters. Second, one of the underlying studies found that some shipments reported as having been pathogen-treated were still contaminated.

Over 37 years, 14 documented outbreaks across nine countries caused 1,946 illnesses, 128 hospitalizations and 2 deaths. Real, and not common.

In practice, spices going into something that will be cooked are not a concern. Spices sprinkled on at the end (on a finished salad, a dip, a garnish) are the case where buying from a supplier you trust matters.

Using them well

Buy whole where you reasonably can. Whole cumin, coriander, peppercorns, cardamom and cinnamon sticks keep their character far longer than ground versions, because grinding exposes surface area to air. A cheap electric grinder or a mortar handles it.

Buy small. Ground spices fade. A jar you will finish in six months is better value than a jar four times the size that will be sawdust by month fourteen, whatever the unit price says.

Store them away from heat and light, which means not in a rack above the stove, however convenient that is. A drawer or a closed cupboard away from the cooker is better.

Test what you have. Open a jar and smell it. If it smells of nothing, it will taste of nothing, and no quantity will fix that. Date the jars when you open them.

Toast whole spices before grinding. A dry pan, medium heat, until they smell like themselves (thirty to sixty seconds). This is the single technique that most changes what a spice contributes.

Bloom ground spices in fat. Add them to the oil early, off the fiercest heat, for fifteen or twenty seconds before the liquid goes in. Ground spices added at the end of cooking taste raw and dusty; the same spices bloomed in oil taste like a dish.

Use more than you think. Home cooks systematically under-season. A teaspoon of cumin in a pot of lentils for four is timid. Two is a starting point.

Buy from the right aisle. Spices sold in small glass jars in the baking aisle are among the most marked-up items in the store. The same spices in the international or bulk section are frequently a quarter of the price, in larger quantities, with faster turnover.

A small set that covers most cooking

If you are starting a spice pantry from nothing, six or seven do most of the work:

Cumin: whole and ground, for beans, lentils, vegetables, anything earthy. Smoked paprika: adds depth to bean dishes that would otherwise need meat. Coriander seed: whole, toasted and ground, bright and citrusy. Black pepper: whole, ground fresh; the pre-ground version is a different and worse ingredient. Dried chili: flakes or a whole dried variety, for heat you can control. Cinnamon: savory as well as sweet, particularly with legumes and squash. Bay leaves: cheap, long-lasting, and the thing that makes a pot of beans taste cooked rather than boiled.

Then fresh herbs when they are available, which are a different ingredient entirely: parsley, cilantro and dill go in at the end and in quantity, and a large handful of parsley in a grain salad is closer to a vegetable than a garnish.