Physical activity at work and during leisure time were studied by using a questionnaire in a random sample of 7495 middle-aged men from the Primary Prevention Study in Göteborg and in 1273 able-bodied male patients with a first myocardial infarction, registered in the Infarction Register in the same city over the period 1968-84. Data on coronary risk factors and socio-economic factors were recorded in the population sample as were data on risk factors and known somatic predictors for prognosis in the infarction group. An inverse and graded association was found between leisure time physical activity and mean diastolic blood pressure, total cholesterol, body mass index, tobacco smoking, socio-economic status and mental stress in the random sample. During the approximate 12-year follow-up, low physical activity during leisure time, but not at work, was associated with an increased risk of coronary deaths and non-fatal infarctions in univariate analysis. Inactive subjects had twice the incidence of total coronary events (9.4%) as physically active contemporaries (4.2%). After controlling for major coronary risk factors, occupational class, diabetes, family history of coronary heart disease and mental stress in a multivariate logistic regression analysis, the association between leisure time physical activity and total coronary events disappeared. Physical activity at work and during leisure time estimated for the 12-month period preceding the first infarction was not associated with long-term prognosis after infarction. Infarction patients assessed to be in need of additional rehabilitation due to somatic restrictions, work-related factors and emotional instability, resumed work later and had a higher mortality and non-fatal recurrence rate during follow-up than patients not considered to require additional rehabilitation. Physical inactivity was not a risk factor for primary and secondary coronary events in this study. The inverse direction of the association between leisure time physical activity and coronary risk factors suggests that increased physical activity alters the risk factor profile in a favourable direction.