To examine the vasospastic cause of myocardial infarction (MI) we studied 1) the incidence of rest angina before MI, 2) clinical features of postinfarction angina and 3) the occurrence of MI in variant angina. 1) Of 178 patients with MI, 60 (34%) experienced rest angina for 1 day to 10 years before the onset of MI. The incidence of rest angina was significantly higher in patients having milder coronary stenosis of 75% or less (15/30, 50%) than in others having severe stenosis of 90% or more (45/148, 30%), p < 0.05. 2) Postinfarction angina with ST elevation was observed in 16 patients (9%) and ST elevation developed in leads with pathological Q waves in all patients. The incidence of postinfarction angina was significantly higher in those having milder coronary stenosis than in others having severe stenosis, (27% versus 5%, p < 0.005). Patients with postinfarction angina experienced rest angina before MI more frequently (81%) than others (29%, p < 0.005). Sublingual nitroglycerin was effective in relieving postinfarction angina attacks and oral calcium antagonist prevented attacks in all patients. 3) MI developed in 9 of 97 patients with variant angina. Six patients had transmural and 3, non-transmural MI. Pathological Q waves and/or coronary T waves appeared in leads where ST elevation was observed during anginal attack. In 7 patients MI developed when antispastic agents were not used and in 2, when angina persisted even under treatment with calcium antagonist. These data strongly suggest that the coronary spasm can be a cause of MI in some patients.