Patients with premenstrual syndrome recorded their symptoms daily using menstrual distress questionnaires. These were analysed by a least mean square method of fitting sine waves. After recording an untreated cycle, patients were given progesterone 200 mg b.d. and placebo in a double-blind crossover manner; 75 per cent of patients were then given progesterone 400 mg b.d. and placebo in a similar manner. Treated cycles were rated by both daily menstrual distress questionnaires and retrospective self-assessment. Both rating methods showed there was no significant difference between progesterone and placebo in reducing symptoms of premenstrual syndrome, and in the majority of cases placebo was more effective, although never significantly so.