Sixty-three patients with stable, severe typical angina pectoris (New York Heart Association functional class III or IV) were treated with propranolol and studied prospectively with a follow-up period of 5 to 8 years to assess the rate of complications and long-term effectiveness after an initial control period. The patients' mean age was 56 years; the mean daily dose of propranolol was 255 mg. The average yearly mortality rate was 3.8 percent with a cumulative 5 year mortality rate of 19 percent. Patients whose reduction of angina with propranolol was less than 50 percent had a nearly four-fold greater mortality rate than those whose reduction was 50 percent or more (P less than 0.01). Thirty-two percent of patients per year were angina-free with propranolol and 84 percent per year had 50 percent or more reduction in anginal episodes. There was no evidence for tachyphylaxis. Heart failure developed in 25 percent of patients, two thirds of whom had either congestive heart failure with an acute infarction or a prior history of congestive heart failure. All patients whose initial cardiothoracic ratio was greater than 0.5 had heart failure during the first 3 years of propranolol therapy. Of 12 patients who had an acute infarction during therapy, 7 died, 6 with cardiogenic shock; in contrast, 8 of 9 patients who had congestive heart failure without acute infarction survived. Eight percent of patients had other significant side effects, including gastrointestinal symptoms (three patients), hallucinations (one) and postural hypotension (one). The occurrence of asthma in three patients was dose-related and did not require drug discontinuation. Propanolol is an effective form of long-term therapy for severe angina pectoris; it does not induce tachyphylaxis or increase the overall mortality rate, although it may increase the risk of cardiogenic shock in acute myocardial infarction. Previous history of congestive heart failure, a cardiothoracic ratio of more than 0.5 without overt heart failure and mild asthma are relative contraindications. A 50 percent or greater reduction in anginal pain with propranolol predicts a low mortality group.