The effect of participation in an employee fitness program on reduction of absenteeism due to illness on three different worksites (police force, chemical industry, banking) was studied. The study used a longitudinal pretest-posttest design. From 884 subjects, data were collected on absenteeism in the pre-intervention year and on absenteeism in the first year of the fitness program. The subjects were divided into three groups: high participation, low participation, and no participation in the employee fitness program. Results of covariance analyses show that the high participation group had a significant decline in sick days (4.8 days), while the low and no participation groups showed no change in sick days. Even when self-selection in the participation in an employee fitness program seems to be present, there is potentially for a great deal to be gained from participation of this self-selected group in terms of changed sick days.