Twelve patients were entered prospectively into a randomized double-blind study comparing the efficacy of nifedipine and isosorbide dinitrate (ISDN) in the treatment of variant angina pectoris due to coronary artery spasm. Using the diary technique, both anginal episodes and nitroglycerin tablets consumed were recorded during the pretrial, no drug period, and both active drug phases. During the baseline pretrial period, an average of 1.1 anginal episodes/day occurred with reduction to 0.28/day during nifedipine treatment and 0.39/day during ISDN treatment. Headache was the major side effect during ISDN treatment, occurring in 9 of 11 (81%) patients; and nonheart failure related pedal edema during nifedipine treatment, occurring in 4 of 12 (33%) patients. Intolerable side effects necessitating cessation of treatment occurred in two patients during nifedipine treatment and in three patients during ISDN treatment. Patients preferred nifedipine over ISDN because of increased efficacy and fewer uncomfortable side effects. We conclude that both nifedipine and ISDN are effective therapy for coronary spasm, but that nifedipine was more effective and was preferred by the majority of patients.