Dispositional compassion has been shown to predict higher well-being and to be associated with lower perceived stress and higher social support. Thus, compassion may be a potential individual factor protecting from job strain. The current study examines (i) whether dispositional compassion predicts job strain and effort-reward imbalance (ERI) or does the predictive relationship run from job strain and ERI to dispositional compassion and (ii) the effect of dispositional compassion on the developmental trajectory of job strain and ERI over a 11-year follow-up. We used data from the Young Finns study (n=723) between 2001 and 2012. The direction of the predictive relationships was analyzed with cross-lagged panel models. Compassion’s effect on the trajectories of job strain, ERI, and their components was examined with multilevel models. First, the cross-lagged panel models demonstrated there was no evidence for the predictive pathways between compassion and job strain or its components. However, the predictive pathways from high dispositional compassion to low ERI and high rewards had better fit to the data than the predictive pathways in the opposite direction. In addition, multilevel models showed that high compassion predicted various job characteristics from early adulthood to middle age (lower job strain and higher job control as well as lower ERI and higher reward). Compassion did not predict job demand/effort. The findings were obtained independently of age, gender, and socioeconomic factors in childhood and adulthood. These findings indicate that compassion may be beneficial in work context. Further, compassion might be useful in the management or prevention of some aspects of strain. Our study provides new insight about the role of compassion in work life.