Diabetic patients with acute myocardial infarction have a poor prognosis, which has been attributed to a higher incidence of congestive heart failure and fatal reinfarction. This study reports on the one-year morbidity and mortality in a randomized study with the aim of testing whether insulin-glucose infusion initiated as soon as possible after onset of myocardial infarction and followed by long-term subcutaneous insulin treatment may have a beneficial effect on outcome in diabetic patients. In all, 306 patients were recruited to the insulin-treated group, while 314 patients served as controls. The overall mortality after one year was 19% in the insulin group compared to 26% among controls (P < 0.05). The treatment effect was most pronounced in patients without prior insulin medication and at low cardiovascular risk. In this stratum the in-hospital mortality was reduced by 58% (P < 0.05) and the one-year mortality by 52% (P < 0.02). The most frequent cause of death in all patients was congestive heart failure (66%), but cardiovascular mortality (congestive heart failure, fatal reinfarction, sudden death and stroke) tended to be decreased in insulin-treated patients. However, this difference did not reach the level of statistical significance. The number of reinfarctions was 53 (28% fatal) in the insulin group compared to 55 (45% fatal) in the control group. The two groups did not differ as regards need for hospital care or coronary revascularization during the year of follow-up. In summary, left ventricular failure and fatal reinfarctions contribute to increased mortality in diabetic patients following acute myocardial infarction. Intensive insulin treatment lowered this mortality during one year of follow-up.