Because of the "healthy worker effect" most occupational cohorts, especially those of persons doing physically heavy work, are selected groups with regard to life expectancy. Health selection is also involved in turnover. Mortality and morbidity were studied on three different exposure levels, defined primarily according to physical demands: heavy level (iron foundries), medium level (manufacture of metal products), and light level (manufacture of electrical devices). The population comprised about 15,700 men who were employed in 1950-1976 in the three branches of the metal industry representing the different levels. The number of person-years of follow-up became about 215,800. During the period 1950-1978, there had occurred 1,407 deaths. Data for the mortality and disability analyses were obtained from the national death and disability registers. Occupational history, morbidity, and reasons for turnover were studied in a sample by means of questionnaire sent to 400 current and 600 former workers from each branch. In addition, a questionnaire concerning occupational history was sent to the next of kin of 450 dead persons. When the three cohorts were compared to the general male population of Finland, the standardized mortality ratio was 124 for the foundry workers, 92 for the metal product workers, and 107 for the electrical workers. The survival curves showed no great differences between the three exposure levels. However, the heavy level had the highest degree of mortality, and the medium level the lowest. Foundry workers showed less social selection than the two other cohorts. Metal product workers seemed to be socially and health selected. Economic fluctuations clearly regulated entries of new employees into the industries and thus introduced additional selective features.