BACKGROUND The long-term effects of leisure-time physical activity (LTPA) on job strain have not been assessed in a large prospective population-based cohort study. AIMS To examine the relationship between the LTPA and the prevalence of job strain. METHODS The participants were 861 full-time employees (406 men and 455 women), aged 24-39 years in 2001, from the ongoing Cardiovascular Risk in Young Finns Study. LTPA was assessed using a self-report questionnaire in 1992 and in 2001. The participants were grouped into four categories according to tertiles of LTPA index at two time points: persistently active, increasingly active, decreasingly active and persistently inactive. Job strain was measured in 2001 by indicators of job demands and job control. RESULTS Baseline LTPA was inversely associated with job strain (P < 0.001) and job demands (P < 0.05) and directly associated with job control (P < 0.05) in both sexes in a model adjusted for the change in 9-year LTPA, age, educational level, occupational status and smoking. Compared with persistently active participants, persistently inactive participants had a 4.0-fold higher job strain after adjustment for the confounders. Similarly, persistently inactive participants had a 2.7-fold higher job demands and a 1.8-fold lower job control. Decreasing physical activity was independently associated with high job strain (P < 0.01) and with low job control (P < 0.01). CONCLUSIONS Participation in regular LTPA during leisure may help young adults to cope with job strain. A long-term benefit of LTPA may play a role in the development of mental well-being.