Why saffron is red, and why it costs what it costs

Saffron is red because of a family of compounds called crocins. They are the pigments in the stigma of the Crocus sativus flower, and they are the reason a few threads will colour a whole pot of rice.

Two other compounds finish the picture. Safranal is what makes saffron smell like saffron. Picrocrocin is what makes it taste faintly bitter. Between them, those three explain almost everything about the spice, including why it costs what it costs and how anyone can tell it apart from an imitation.

Crocins: the colour

Crocins are carotenoid pigments, related to the compounds that make carrots orange, but with an important difference. Most carotenoids will not dissolve in water. Crocins will, which is why saffron colours a broth all the way through instead of floating on the surface.

They are not a single substance. Up to 15 different crocin isomers have been identified in one standardised saffron extract. The major one, known as trans-4-GG, made up between 36.1 and 64.7 per cent of the total crocins across the extracts analysed in a 2023 study, with trans-3-Gg the next most abundant at 8.6 to 23.8 per cent. (Mena-Garcia et al., Foods, 2023;12:1634.)

The proportions of those isomers are stable enough to act as a fingerprint. Two saffron samples from different farms will have broadly the same pattern, and a powder pretending to be saffron generally will not.

Safranal: the smell

Safranal is a volatile aroma compound, and there is remarkably little of it. In the standardised extract used in most of the published trials it accounts for 0.03 per cent, which is three hundredths of one per cent by weight.

That tiny figure does an enormous amount of sensory work. It is why saffron is recognisable across a kitchen and why the smell fades from an old jar long before the colour does.

Picrocrocin: the taste, and the proof

Picrocrocin is the bitter one, and it is also the compound that authenticates saffron. Raw stigmas measured 89 to 119.9 milligrams of picrocrocin per gram in the 2023 analysis. Gardenia fruit, which is the cheap material most often used to imitate saffron powder, contains crocins but no picrocrocin at all, so it is the compound that separates the real thing from a convincing orange substitute.

Safranal and picrocrocin are chemically related. Picrocrocin breaks down during drying and storage, and safranal is one of the things it becomes, which is why saffron develops much of its characteristic aroma after harvest rather than arriving with it.

Which compound carries the effects

Researchers who study the extract have begun to answer this. The authors of a 2026 trial concluded that the results they observed were most likely driven by the crocin fraction, with only a limited contribution from safranal, on the reasonable grounds that a standardised extract contains far more crocin than safranal. (Lopresti and Smith, Frontiers in Nutrition, 2026;13:1838513.)

That is one team's reading of their own data rather than a settled matter, but it points at why a standardised extract is measured on crocins first.

Why it costs what it costs

Each crocus flower produces exactly three stigmas, and they are removed by hand, flower by flower.

It takes somewhere between 90,000 and 170,000 flowers to make one kilogram of dried stigmas. There is no mechanical harvest, no shortcut and no variety that yields more than three. That arithmetic, and not scarcity of land, is why saffron has been the most expensive spice in the world by weight for as long as anyone has kept records.

Why the colour is not identical every time

One last thing that surprises people. A specification sheet for a saffron extract will often carry a note that natural variation in the raw material can produce colour differences from batch to batch without affecting the quality of the product.

That is honest, and it is the correct way round. An agricultural ingredient that looked exactly the same every single time would be more suspicious, not less. It is why a serious extract is judged on a measured percentage of named compounds rather than on how it looks in the tub.

Studies conducted on individual ingredients, not on any finished product. Individual results may vary.