In this paper some previously found associations between psychosocial occupational characteristics and myocardial infarction (MI) risk are scrutinized regarding confounding effects. Standardized occupational characteristics were obtained for 118 occupational groups by means of a nation wide interview survey (3876 men). Possible confounding factors available were smoking, low level of education, high proportion of immigrants (mainly of Finnish origin) and heavy lifting. The standardized characteristics were utilized in a case-control study of 1216 men 40-64 years of age, living in the Stockholm County. It was concluded that occupations characterized by both high demand and at the same time small possibility of control or growth ('strain') are associated with an elevated MI risk regardless of the confounding factors. Taking each of the confounding factors into account the risk of developing MI for men aged 40-54 years employed in these strenuous occupations is about twice as high as for those employed in other occupations.