The cyclical exacerbations of epilepsy (catamenial epilepsy) were used to assess the antiepileptic effect of a benzodiazepine, clobazam. Doses of 20 mg, and in some cases 30 mg, per day were compared with placebo over predetermined ten-day periods in a double-blind cross-over study. The results were evaluated by preference in a sequential procedure. In 14 of 18 patients who received both treatments clobazam was superior to placebo, and in 4 patients no preference was established. Clobazam completely prevented seizures in most of the patients, and toxic effects were of low frequency and severity.