We conducted a randomized open-labeled study of nifedipine versus propranolol for the initial prophylaxis of migraine. Propranolol was effective in 67% of patients (12/18) and well tolerated. Nifedipine was effective in only 30% of patients (6/20). The lack of overall efficacy of nifedipine was attributable to a high incidence of side effects, including an unusual symptom complex resembling erythromelalgia. These side effects led 45% (9/20) of the nifedipine patients to withdraw from the study within two weeks. By contrast, no patient (0/18) withdrew from the study within the first 2 weeks of propranolol therapy. We conclude that nifedipine is not an agent of first choice for the prophylaxis of migraine.