Seventy-nine men who had suffered a myocardial infarction before the age of 45 while they were vocationally active in the greater Stockholm area were followed for five years. Forty-nine survived without cardiac complications and 13 died due to ischaemic heart disease during the period of follow-up. These two contrasting groups were compared with regard to psychosocial risk factors at work before the first myocardial infarction (as reported by the patient when he was interviewed during the weeks after the onset of disease). It was hypothesized that returning to stressful work (high demands and limited possibilities of influencing decisions and developing skills) would be associated with an increased risk of death. All the subjects who died had returned to the work that they had performed prior to the first myocardial infarct. Work performed by these patients was described as having significantly higher psychological demands in relation to the possibility of learning new things and higher demands in relation to variety as well as almost significantly higher demands in relation to influence. Multivariate logistic regression with these factors concerning employment, together with biomedical risk factors recorded at the same time, showed that increasing age, increasing degree of coronary atherosclerosis and number of stenosed coronary arteries, as well as high demands in relation to the possibility of learning new things, were independent predictors of death due to coronary arterial disease. The remaining 17 subjects either survived a re-infarction, or had coronary arterial by-pass surgery during the period of follow-up. This heterogeneous group occupied intermediate positions with regard to psychosocial job factors.