AIMS: The present study was performed to evaluate pre-admission history, presentation, initial treatment and long-term mortality in patients with myocardial infarction and diabetes. METHODS AND RESULTS: Between 1990 and 1992, 6676 patients with acute myocardial infarction were screened for entry into the Trandolapril Cardiac Evaluation (TRACE) study. In this cohort 719 (11%) of the patients had a history of diabetes. Among the diabetic patients 19% were treated with insulin, 52% with oral hypoglycaemic agents and 29% with diet only. The diabetic patients were slightly older, more likely to be female and had a higher prevalence of known cardiovascular disease. Even though the diabetic patients had the same frequency of ST-segment elevation on the electrocardiogram and the same admission delay, treatment with thrombolysis and aspirin was less frequently prescribed to the diabetic patients than to patients without diabetes. The mortality rate was significantly increased in the diabetic patients, 7-year mortality being 79% in insulin-treated, 73% in tablet-treated and 62% in diet-treated diabetic patients compared with 46% in patients without diabetes. In a multivariate analysis only diabetic patients treated with oral hypoglycaemic agents or with insulin had an increased mortality compared with non-diabetic patients. CONCLUSIONS: Patients with diabetes mellitus and myocardial infarction are treated with thrombolysis to a lesser extent than non-diabetic patients. Diabetic patients treated with oral hypoglycaemic agents or insulin, but not those treated with diet alone, have a significantly increased mortality following acute myocardial infarction compared with non-diabetic patients.