Aims/hypothesis. Mortality of diabetic patients after myocardial infarction remains high despite recent improvement in their management. This study population-based evaluates the impact of cardiovascular drug therapy on mortality within 28 days and during 5-year follow-up in diabetic compared with non-diabetic patients.¶Methods. Using the MONICA Augsburg register from 1985 to 1992, 2210 inpatients with incident Q-wave myocardial infarction aged 25–74 years were included, of whom 468 had diabetes. Primary end point was mortality within 28 days and over 5 years. General linear model procedures were used for age-adjustment, controlling for sex, and testing significance; hazard risk ratios were calculated using multivariable Cox proportional hazards model procedures.¶Results. During the 5-year follow-up, 598 subjects died (396 diabetic, 202 non-diabetic). The mortality rate within 28 days was 12.6 % in diabetic patients (women 18.0 %, men 9.9 %) and 7.3 % in non-diabetic patients (p = 0.001). Mortality in diabetic patients over 5 years was increased by 64 % (95 % confidence interval 1.39–1.95) compared with non-diabetic patients. This was considerably reduced (p < 0.001) in patients treated with thrombolytic drugs (risk ratio: diabetes 0.57, no diabetes 0.65) and with beta blockers (0.62 and 0.64) and antiplatelets (0.76 and 0.74) at hospital discharge. Mortality of diabetic patients treated with these drugs was reduced to that of non-diabetic patients without such treatment (risk ratio 1.01 to 1.27; p > 0.1).¶Conclusion/interpretation. Diabetic patients after myocardial infarction are at particularly high risk of dying, but benefit clearly from treatment with thrombolytics, beta blockers and antiplatelets. This study does not, however, allow any inferences to be drawn for treatment with angiotensin converting enzyme inhibitors or the impact of left ventricular function. [Diabetologia (2000) 43: 218–226]