The role of coronary angiography before hospital discharge after myocardial infarction was assessed in 1043 hospital survivors of the alteplase/placebo and the alteplase/PTCA trial of the European Cooperative Study Group. Forty-two of 1043 patients (4.0%) died after 1 to 489 days after predischarge coronary angiography. In survivors, follow-up ranged from 34 to 1106 days. In a stepwise multivariate regression model (Cox), use of diuretics and/or digitalis, a history of previous infarction and age exceeding 60 years were retained in the model with clinical data only. In addition, inability to perform exercise testing and less than 30 mmHg exercise-induced systolic blood pressure increase were selected by multivariate analysis. Large enzymatic infarct size, radionuclide left ventricular ejection fraction below 40%, and multivessel disease were also determinants of mortality after hospital discharge. The risk function, including coronary angiography, performed no better in late mortality prediction than functions based on clinical data and non-invasive testing. Patients without a history of previous infarction, not treated with diuretics and/or digitalis and with a systolic blood pressure increase of 30 mmHg or more during exercise had an excellent survival (98.6%) in the first year after hospital discharge, irrespective of whether symptoms of recurrent ischaemia occurred. This low risk group formed 47% of the total patient population and does not benefit from coronary angiography.