In a clinical trial, methysergide and placebo were compared by the double‐blind method. The study was arranged so that it was also possible to assess the effect of methysergide and that of placebo in relation to the pre‐experimental condition. A dose of 6 mg. of methysergide was given to adult patients daily. Sixty patients completed the trial satisfactorily. Analysis showed that there was significant reduction in the frequency of attacks with methysergide. In comparison with the pre‐experimental period the frequency of attacks was reduced at least 50 per cent in 57 per cent of the patients by methysergide. The corresponding figure for placebo was 27 per cent, but part of this response seemed to be due to a “carry‐over” effect. No definite correlation between the effect and the clinical data was revealed. Severe adverse effects necessitating the withdrawal of the drug occurred in 7 per cent of patients; other untoward effects were observed in about one third of the patients, mainly due to insufficient circulation. Adverse effects were more frequent in women, and especially in those who obtained good effects from the drug.