Data previously collected by the Danish Institute of Clinical Epidemiology (DICE) in 1986/87 by a structured interview of a sample of the adult Danish population was analysed to evaluate sick leave among 1) persons with preschool children, 2) persons with either physically demanding work or with exposure to chemicals, and 3) persons with mentally demanding employment. Average absence from work in the year previous to the interview was increased in unmarried women with preschool children (9.8 days). Average absence from work increased in a stepwise fashion from 4.7 to 18.7 days with increasing physical work loads, and from 5.6 to 10.1 days with increasing exposure to chemicals. In the group with mentally demanding work, the increased was from 6.3 to 20.0 days up to a load level of five. Risk of long lasting (above 10 days) sick leave was increased among the 52.9% of persons in the study base who either had physically demanding work or who were exposed to chemicals, prevalence rate 1.91 (1.56-2.35, 95% CI), as well as among the 20.9% of persons in the study base with mentally demanding work, prevalence rate 1.68 (1.38-2.05, 95% CI). Logistic regression analysis revealed statistically significant odds ratios also when age, tobacco smoking, marital status, chronic disease and overlap between the risk groups were taken into consideration.