The aim of the study was to investigate whether workers in jobs dominated by the opposite sex have an increased risk of myocardial infarction (MI). A case-referent study was carried out to estimate the relative risk of first MI in different occupational groups. The study base comprised all men and women in five counties in the middle of Sweden during 1976-84. Cases of MI were identified from both hospital discharge records and death records. Information on occupation was obtained from two consecutive censuses. Primary health related selection was analysed for men with data from the physical examination of conscripts to compulsory military service in 1969-70 combined with data from the censuses of 1970-90 and data on early retirement in 1971-92. Increased risk of MI was found among both women (relative risk (RR) 1.41, 95% confidence interval (95% CI) 1.15 to 1.73) and men (1.21, 1.10 to 1.32) in blue collar jobs where men predominate, and among men with white collar jobs (1.26, 1.09 to 1.45) where women predominate. However, the increased risk among men in white collar jobs was probably due to negative health selection into these occupations. These results do not support the notion that being of the sexual minority in an occupation is in itself an important risk factor for MI.