Saffron for Depression and Anxiety: What 20 Years of Trials Show, and What They Do Not

Saffron is the only kitchen spice with a serious body of randomised trials for depression, and the results are better than most people expect: in mild to moderate depression, 30 mg a day has repeatedly beaten placebo and matched low doses of standard antidepressants over six to eight weeks. The results are also less than enthusiasts claim: the trials are small, short, largely from a few research groups, and silent on severe depression and on anything longer than three months.

This guide lays out the evidence honestly, explains the dose and what to expect, and is clear about who should not try it. It summarises published research and is not medical advice. Depression is a medical condition; if you have it, see a clinician, and do not change prescribed treatment on the strength of an article. Our disclaimer explains how we handle health topics.

The trials, briefly

The modern research began in Tehran in the early 2000s. A series of small, double-blind trials from Shahid Beheshti and Tehran universities compared 30 mg a day of saffron stigma or petal with placebo, with fluoxetine 20 mg or with imipramine 100 mg, in adults with mild to moderate major depression, over six weeks. Saffron beat placebo and was statistically indistinguishable from the antidepressants on standard depression scales, with side effects no worse and in some comparisons milder.

Over the following decade trials from Iran, then Australia and Europe, repeated the result with dried stigma, with standardised extracts and in different groups: adults with mild depression, people with depression and anxiety together, adolescents with subclinical symptoms, postpartum women, and people with depression alongside type 2 diabetes or coronary disease.

Several meta-analyses have pooled them. A 2013 analysis in the Journal of Integrative Medicine, a 2014 review in Human Psychopharmacology, a 2019 meta-analysis in Planta Medica and a 2020 analysis in the Journal of Nervous and Mental Disease all reached the same shape of conclusion: saffron produces a large effect versus placebo and a null difference versus antidepressants, from small trials of mostly moderate quality, with a need for larger and longer studies. None found a safety signal at 30 mg.

What the evidence does not cover

Honesty about the gaps matters more here than anywhere else on this site.

Severity. Almost all trials enrolled mild to moderate depression. There is no evidence for saffron in severe depression, in depression with psychotic features or in people at risk of self-harm, and nobody should treat those with a spice.

Duration. Six to twelve weeks. Depression is often a relapsing condition and antidepressants are typically taken for months to years. Whether saffron’s benefit holds, fades or has long-term effects is unknown.

Size and independence. Most trials had 40 to 80 participants, and a large share came from a small number of groups, some with ties to saffron producers or extract makers. Independent replication at scale has not happened.

Comparators. Saffron matched fluoxetine at 20 mg and imipramine at 100 mg, the low end of therapeutic doses. It has not been compared with optimised antidepressant treatment or with psychotherapy.

How it might work

Laboratory work suggests saffron’s compounds, crocin and safranal in particular, inhibit serotonin reuptake, modulate dopamine and noradrenaline, reduce inflammation and oxidative stress, and may affect the stress-hormone axis. Antidepressants target some of the same systems. This fits the clinical findings but has not been shown to be the mechanism in people.

Anxiety

The anxiety evidence is thinner and more mixed. Trials in adults with mixed anxiety and depression, and one in adolescents using 14 mg twice a day, reported reductions in anxiety scores; a trial in generalised anxiety was small. A 2019 meta-analysis found a moderate effect on anxiety symptoms with the same caveats. Taken together, saffron may help mild anxiety, especially where it travels with low mood, and there is no good evidence for anxiety disorders on their own.

Dose, form and timing

The trial dose is 30 mg a day of saffron stigma or an equivalent standardised extract, usually as 15 mg twice a day or 30 mg once. In threads, 30 mg is about 12 to 15 of good Kashmiri Mongra, steeped for ten minutes in warm milk or water and drunk, threads and all. Saffron dosage per day has the conversions.

Expect nothing for the first two weeks, possible change by week four, and judge at week eight. Side effects at this dose are mild and uncommon: occasionally nausea, dry mouth, headache or drowsiness. Do not exceed the dose; more has not been shown to work better, and above 1.5 g a day saffron becomes unsafe. Saffron side effects sets out the full picture.

Who should not try it alone

Anyone with moderate to severe depression, or thoughts of self-harm. See a doctor. Saffron is not studied in your situation.

Anyone on an antidepressant. Adding a serotonergic supplement to an SSRI or SNRI should be a decision made with the prescriber. Never stop or taper a prescribed antidepressant to try saffron.

People with bipolar disorder. Case reports describe mania after saffron supplements, consistent with its mood-elevating action. Avoid, or use only under psychiatric supervision.

Pregnant or breastfeeding women. Culinary amounts only; the postpartum trials used 30 mg but under medical oversight.

Anyone on blood thinners or blood pressure medication. Saffron has mild antiplatelet and blood-pressure-lowering effects; check with your doctor.

If you decide to try it

Use real saffron. A daily 30 mg of dyed safflower, which is what much cheap saffron is, has no crocin and no safranal in it. Buy whole threads with a per-lot ISO 3632 report and run the water test: a slow yellow over 15 minutes while the threads stay red. How to spot fake saffron has the tests. Our lots are tested one by one and the number is on the vial; see the saffron page. At 15 threads a day, a 1 g vial lasts a month.

Keep whatever else helps: sleep, movement, daylight, people, and treatment if you need it. Saffron is a small, pleasant addition to that list, not a replacement for any item on it.

Read next: Saffron benefits: what the research actually says and Saffron for sleep.

Questions people ask

Does saffron help with depression? +

In mild to moderate depression, a dozen or so small randomised trials have found 30 mg a day of saffron better than placebo and roughly comparable to low doses of fluoxetine or imipramine over six to eight weeks. Meta-analyses agree the effect is real but note that the trials are small, short, and mostly from a few research groups. Saffron has not been tested in severe depression and is not a replacement for treatment.

Does saffron help with anxiety? +

The anxiety evidence is thinner than for depression. Several trials, including some in adolescents and in people with both low mood and anxiety, reported reduced anxiety scores on 28 to 30 mg a day over eight weeks. A 2019 meta-analysis found a moderate effect on anxiety symptoms, with the same caveats about size and quality.

How long does saffron take to work for mood? +

Trials measured improvement at six to eight weeks, with separation from placebo often visible by week four. Nobody should expect a change within days. If nothing has shifted after eight weeks at 30 mg a day, it is unlikely to.

Can I take saffron with my antidepressant? +

Only with your prescriber's agreement. Saffron appears to act partly through serotonin, and a few trials added it to an SSRI with good tolerability, but combining serotonergic agents is something to do under supervision, not on your own. Do not stop or reduce a prescribed antidepressant to try saffron.

Which is better for mood, saffron threads or extract capsules? +

Trials used both dried stigma in capsules and standardised extracts, at 30 mg a day, with similar results. Fifteen threads of lab-tested Category I saffron steeped in milk delivers an equivalent amount. Capsules offer precision; threads offer certainty about what you are taking, provided the lot has a lab report.