Whether diabetic patients may benefit, compared with non-diabetic patients, from beta-blocker therapy following acute myocardial infarction was examined in a large multicentre cohort of 2024 patients, including 340 diabetics, 281 of whom survived hospitalization. One-year mortality following discharge was 17% for diabetics compared with 10% for non-diabetics (P less than 0.001). However, diabetics discharged on beta-blockers had a 1-year mortality of only 10%, compared with 23% for diabetics not on beta-blockers. In non-diabetics, mortality rates were 7% and 13% for those taking and not taking beta-blockers, respectively. Bias in patient selection for beta-blocker therapy might be responsible for the trends exhibited in our population since patients were not randomized to treatment. In diabetics, evidence of pulmonary congestion on X-ray was more prevalent than in non-diabetics; this appeared to be true both for patients taking beta-blockers and for those not taking beta-blockers. However, even in diabetics without evidence of pulmonary congestion on X-ray, 1-year mortality was 7% vs 17% for those with and without beta-blocker therapy, respectively (P less than 0.04). In multivariate analysis, beta-blocker use was an independent predictor of 1-year cardiac survival following hospital discharge for all diabetics, even those without evidence for pulmonary congestion on X-ray, but not for non-diabetics. These data suggest a beneficial effect, but a definitive answer regarding the benefit of beta-blockade in diabetic patients after acute myocardial infarction would require a prospective, randomized study.