Homoeopathic remedies for migraine are widely available over the counter, statutorily offered by the national health service in the UK, and apparently popular with patients. Do they work? Sixty-three outpatients with migraine with or without aura by IHS criteria entered a 4-month randomized placebo-controlled, double-blind parallel-groups trial of individualized homoeopathic prophylaxis, the first month being baseline with all patients on placebo. Three patients (4.8%) dropped out, leaving 30 in each treatment group. There were chance differences in attack frequency and severity between the groups at baseline (attacks were more frequent but less severe in the placebo group). Both groups improved on therapy, but neither to a great extent on the primary outcome measure of attack frequency (verum: -19%; placebo: -16%). Reduction was mostly in mild attacks on placebo, more in moderate and severe attacks on homoeopathy. Few adverse events were reported. Overall, there was no significant benefit over placebo of homoeopathic treatment. The course of change differed between groups, and suggested that improvement reversed in the last month of treatment on placebo. On this evidence we cannot recommend homoeopathy for migraine prophylaxis, but cannot conclude that it is without effect.