How much saffron the research actually used

The amount used in the research is 28 milligrams a day of a standardised saffron extract, taken as 14 milligrams in the morning and 14 milligrams in the evening.

That figure is specific rather than approximate, and there is a good reason for it. One trial tested a slightly smaller amount alongside it, and the smaller amount did nothing.

The trial that tested a lower amount

In 2017 Kell and colleagues ran a four-week randomised, double-blind, placebo-controlled trial in 128 adults who reported low mood but had no diagnosis. It had three arms: 28 mg of the standardised extract a day, 22 mg a day, and placebo. (Complementary Therapies in Medicine, 2017;33:58-64, PMID 28735826.)

At 28 mg, total mood disturbance scores on the Profile of Mood States improved against placebo (p < 0.001, d = 1.10). At 22 mg there was no treatment effect at all.

Six milligrams separated a clear result from nothing measurable. That is the most useful single fact in this literature for anybody reading a label, because it means the word saffron on an ingredients list tells you almost nothing on its own. The number next to it is the entire question.

Where the morning and evening split comes from

Three trials have used the same 14 plus 14 arrangement, and they are the three most relevant ones.

• 80 young people aged 12 to 16, eight weeks. 14 mg twice a day or placebo. 68 completed. (Lopresti et al., Journal of Affective Disorders, 2018, PMID 29510352.)
• 202 adults, twelve weeks. One 14 mg tablet in the morning and one in the evening, delivering 28 mg a day. This is the largest trial of the extract published. (Journal of Nutrition, 2025, PMID 40414301.)
• 86 women aged 50 to 70, twelve weeks. The same morning and evening split. (Lopresti and Smith, Frontiers in Nutrition, 2026;13:1838513.)

Splitting the day is not a stylistic choice in these protocols. It is what was tested, and it is the arrangement the results belong to.

With food or without

Both of the two largest trials specify the point explicitly in their methods: the tablets were taken in the morning and in the evening, with or without food.

That answers a question people ask constantly about supplements, and for this ingredient the answer is on the record rather than inferred. It does not have to be attached to a meal.

Milligrams of what, exactly

One trap is worth naming, because it makes label comparisons meaningless if you miss it.

The 14 mg refers to the standardised extract itself. It does not refer to a marker compound inside the extract. Crocins make up around three and a half per cent of a standardised extract, so a crocin figure describing the same material is a much smaller number, and the two cannot be read side by side.

This is not hypothetical. Two Japanese trials of this extract stated their amounts in milligrams of crocin rather than milligrams of extract, and reconciling them with the Australian trials means working the extract amount back out from the standardisation percentage. When you are comparing two products, check that both numbers describe the same thing before you conclude that one is stronger.

How long the trials ran

Four weeks at the shortest and twelve at the longest, with both of the largest trials running the full twelve weeks and the adolescent trial running eight.

Nothing in this literature was measured over a few days. If you are judging an ingredient against the research, the research gave it two to three months.

One honest note on age

The youngest person enrolled in any placebo-controlled trial of this extract is 12 years old. The 14 plus 14 split is what that trial used and what the two largest adult trials used, but there is no trial of this extract in children under 12, and nobody should imply otherwise.

What to take from it

Twenty-eight milligrams a day of a standardised extract, split into morning and evening, with or without food, over eight to twelve weeks. That is the shape of the evidence. A product that carries a fraction of that amount has borrowed the reputation of this research without the thing that produced it.

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