We examined the role of chronic (greater than 1 month) angina pectoris (AP) before acute myocardial infarction (AMI) in predicting hospital and long-term mortality rates among 4166 patients with first AMIs. The prevalence of AP in these patients was 43%. Chronic AP was more common in women (49%), patients with hypertension (49%), and diabetic patients (49%) than in men and counterparts free of the former conditions (p less than 0.005). In patients with AP the hospital course was more complicated and non-Q-wave AMI was more common than in counterparts without AP. In-hospital (16%), as well as 1 (8%)- and 5-year postdischarge (26%), mortality rates in hospital survivors were higher among patients with previous AP than in patients without previous AP (12%, 6%, and 19%, respectively) (p less than 0.01). After adjustment for age and all other predictors of increased hospital mortality rates in this cohort of patients, AP preceding AMI emerged as an independent predictor of increased hospital mortality rates (odds ratio 1.30; 90% confidence interval 1.10 to 1.53). For postdischarge mortality rates (mean follow-up 5 1/2 years), the covariate-adjusted relative risk of death in patients with AP was similar at 1.29 (p less than 0.0001; 90% confidence interval 1.16 to 1.44), according to estimation by Cox proportional hazards model. These data support the notion that preexisting AP identifies a group of patients at increased risk of death.