Our buying criteria for saffron are six short lines. This is the working behind them — where the numbers on a saffron label come from, which of them are real, and one that is not. You do not need any of it to buy saffron. It is here because we would rather show it than ask you to take our word.
Whether what is in the packet is saffron at all, and whether it still carries its compounds. Saffron is the most valuable spice on earth by weight, which is exactly what makes it the most worth faking. Everything below is a way of answering that one question.
Every mood trial we could read used an extract, not threads. Thirty milligrams a day, six weeks: in the 2004 comparison against a standard antidepressant, as three ten-milligram capsules; in the SaffroMood trials, as two fifteen-milligram capsules. Not one of them used loose saffron. That is the fact our buying page has to deal with honestly, because what we tell you to buy is threads.
One extract publishes the arithmetic. affron states a three-to-one plant-to-extract ratio, and its maker's own published filing gives a stigma equivalence of 90 milligrams for a 30 milligram capsule. So for that one product, a clinical dose is about 90 milligrams of real saffron. Now the limit: SaffroMood publishes no ratio at all, and the 2004 trial describes its method as 120 grams of dried and milled stigmas extracted with ethanol — which works out closer to twelve to one. There is one product's conversion here. There is no field-wide one, and anybody printing a general threads-to-capsule rule has invented it.
Saffron extracts are widely sold as standardized to 2 percent safranal. Real saffron runs about 0.1 to 0.2 percent safranal. The figure comes from taking the measurement method built for the raw spice and applying it to a concentrate, where it does not belong, and extracts carrying that label have been found with no detectable safranal in them at all. For contrast, here are affron's own measured numbers: safranal 0.04 percent, total crocins 5.33 percent. If a label leads with 2 percent safranal, it is quoting a number that the plant does not produce.
Super Negin, Negin, Sargol and Pushal describe how the stigma was cut. Super Negin is the whole red stigma. Sargol is the trimmed red tips, which is also the most broken. Pushal keeps one to three millimeters of yellow style attached — weight that carries no color, no scent and no taste. No standards body defines any of these words. What can be measured is the ISO 3632 category, which grades saffron on coloring strength. Sellers routinely claim the grade names map onto the categories. They do not reliably, because a category is a measurement of one batch and a grade is a description of a cut. This is why our criterion asks for both.
A European survey of nearly 1,900 herb and spice samples put saffron among the most adulterated spices in the food supply. A study of 80 retail samples from one growing region found only 40 percent met ISO Category I, with adulteration running to around 40 percent of samples in some districts. What it is cut with: dyed corn silk, dyed styles from the same flower, dyed paper strips, safflower and turmeric.
The three compounds do not come out under the same conditions. Work published in 2024 found crocin, the color, extracts best at 60 degrees over 20 minutes in water, and degrades if pushed hotter. Picrocrocin, the taste, wants near-boiling. Safranal, the scent, favors a medium containing fat and breaks down fastest at 100 degrees. So the traditional preparation — saffron bloomed in warm milk rather than dropped in boiling water — is not sentiment. It is the condition that suits the most fragile of the three. A quick dunk in hot water gets you color and very little else.
No trial has tested loose saffron threads brewed as a tea. The closest anyone has come is a water extract and a powdered-stigma tablet. The chemistry does transfer, and one maker's own filing supports the arithmetic, but nobody has measured a cup of saffron tea against a placebo — and they would struggle to, because you cannot blind anyone to a drink that turns bright yellow and smells like saffron. So we recommend the threads because that is what we would buy, not because a trial says so. When we have a better reason we will put it here.
Because the buying section is how we chose, not homework for you. The page says to check the species, to want the red stigma with a category named, to look for deep color and a warm scent, to avoid hidden fillers, to want an origin, and to want it tested. This is what we do when we apply that.
We have no product of our own and take no payment for a place on a list. Where a figure below is not ours, the source is named so you can check it yourself.