The incidence of congestive heart failure was studied in the Beta Blocker Heart Attack Trial in which postmyocardial infarction patients between the ages of 30 and 69 years, with no contraindication to propranolol, were randomly assigned to receive placebo (n = 1921) or propranolol 180 or 240 mg daily (n = 1916) 5 to 21 days after admission to the hospital for the event. Survivors of acute myocardial infarction with compensated or mild congestive heart failure, including those on digitalis and diuretics, were included in the study. A history of congestive heart failure before randomization characterized 710 (18.5%) patients: 345 (18.0%) in the propranolol group and 365 (19.0%) in the placebo group. The incidence of definite congestive heart failure after randomization and during the study was 6.7% in both groups. In patients with a history of congestive heart failure before randomization, 51 of 345 (14.8%) in the propranolol group and 46 of 365 (12.6%) in the placebo group developed congestive heart failure during an average 25 month follow-up. In the patients with no history of congestive heart failure, 5% in the propranolol group developed congestive heart failure and 5.3% in the placebo group developed congestive heart failure. Baseline characteristics predictive of the occurrence of congestive heart failure by multivariate analyses included an increased cardiothoracic ratio, diabetes, increased heart rate, low baseline weight, prior myocardial infarction, age, and more than 10 ventricular premature beats per hour. Patients with congestive heart failure in the propranolol group experienced a similar decrease in the total mortality (27%) compared with those without congestive heart failure (25%), whereas propranolol decreased the occurrence of sudden death by 47% in the patients with prior heart failure compared with 13% without congestive heart failure. Circulation 73, No. 3, 503-510, 1986. THE USE OF ,3-adrenergic blocking agents in survivors of acute myocardial infarction with evidence of congestive heart failure poses a therapeutic paradox for the clinician. A number of 3-adrenergic blocking agents have been shown to decrease the mortality rate in postmyocardial infarction patients when administered between 1 and 3 weeks after the event. 1-' Patients with evidence of severe heart failure were usually but not always excluded from enrollment in these studies. These reports have been supplemented by both prospective and retrospective studies suggesting that a relatively greater benefit, including a significant reduction in sudden death, may be realized in the group of patients at the highest risk of future events. The Beta Blocker Heart Attack Trial (BHAT) offers an opportunity to evaluate the effect of long-term therapy with propranolol in patients after an acute myocardial infarction with congestive heart failure. As part of the design and safety features, patients with shock or overt heart failure were excluded from enrollment in BHAT until these conditions were ameliorated. In this analysis we wi...