Aims/hypothesis. Diabetes mellitus is associated with increased mortality in subjects with acute myocardial infarction (AMI). We aimed to estimate the risk of mortality in AMI patients with and without diabetes using the urinary albumin : creatinine ratio (ACR).Methods. This is a prospective study of 121 consecutive, non-selected diabetic AMI patients, 121 age-and sex-matched non-diabetic AMI patients and 61 diabetic non-AMI outpatients as control subjects. All data were obtained during the first 7 days of hospitalisation and each AMI patient was followed for a period of exactly 3 years. Baseline ACR RIA measurements were made on the 1st, 3rd and 7th days of admission. Results. Adjusted ACR values were significantly higher in the diabetic AMI patients than in the diabetic control outpatients (p<0.0001), and a significant difference was observed between the weekly ACR slopes for these two groups (p<0.0001). Microalbuminuria was more prevalent in the diabetic AMI patients than in the non-diabetic AMI patients on the 1st day (62% vs 46%, p=0.01) and 3rd day (41% vs 29%, p=0.04). Among the AMI patients with normoalbuminuria (ACR <30 µg/mg), the mortality rate was 11.6% for the patients without diabetes and 33.8% for those with diabetes (p=0.001). The mortality rate was much higher among the AMI patients with microalbuminuria (ACR ≥30 µg/mg) and similar for the diabetic (68.0%) and non-diabetic patients (74.3%). In a multivariable Cox model, ACR (p<0.0001) and diabetes status (p=0.01) were associated with adverse outcome even when several other clinical variables were included in the model. Furthermore, a negative interaction was found between diabetes and ACR (p=0.01). Conclusions/interpretation. Microalbuminuria frequently occurs in diabetic and non-diabetic AMI patients during the first 3 days of admission to hospital and can be used to identify subjects at high risk of mortality.