Coronary morphology, risk factors, long-term prognosis, and progression of coronary arteriosclerosis were investigated in 679 (649 mean and 30 women) post-infarction patients under 40 years of age. These patients represented 80% of 844 MI patients under 40 who were referred to our hospital in the years 1973-1980; 20% had refused coronary angiography; 465 patients were followed up for 1-7 years (mean 3.5 years). In 164 patients, a second coronary angiography was performed 3.8 years after the first angiogram, which was done an average 3 months after the acute episode. The main results were as follows: 8.4% of the patients had zero-vessel disease and 3.7% had a normal coronary angiogram. The majority had single-vessel disease (57.3%). The prevalence of zero-vessel disease decreased with age while that of multivessel disease increased. With increasing vessel involvement, the prevalence of hypercholesterolemia, hypertriglyceridemia, and hypertension increased. A history of smoking was equally common in patients with zero-, single-, double-, and triple-vessel disease. In women the combination of smoking and the use of oral contraceptive drugs was frequently seen. In one-quarter of the zero-vessel disease patients, the infarction occurred during unusually intense physical exercise. The statistical analysis of the survival data using the proportional hazards model (univariate analysis) showed the variables heart volume/body wt., ventricular arrhythmias, PCP at rest, PCP max, work capacity, ventricular function, and number of diseased vessels to be of prognostic importance. Multivariate analysis using this model revealed the following independent variables to be relevant to prognosis: heart volume/body wt., ventricular arrhythmias, ventricular function, and number of diseased vessels. After an average of 3.8 years since the first coronary angiography, 28.6% of the patients showed a significant progression of coronary arteriosclerosis (at least two degrees of stenosis according to the AHA classification). In the subgroup of patients with multilocular disease in the first angiogram, progression was 10 times as frequent as in a group with initial unilocular disease (34.3% vs 3.6%). Patients with progression had continued to smoke significantly more often than patients without progression (38.4%) vs 14.5%). Regression of coronary angiographic findings was significantly more frequent in the group of patients with initial unilocular disease than in those with multilocular disease in the first angiogram (28.6% vs 10.6%). Regression might be explained as recanalization and organization of a thrombus.(ABSTRACT TRUNCATED AT 400 WORDS)