Abstract
Karasek'sJob Strain model was tested in a sample of 122 office workers. Two rounds ofdata were collected and used to examine the reliability of the results and the relationships between job characteristics and worker strain over time. Moderated regression analyses based on standard scores of the independent variables were conducted to examine the interaction of job control and workload, and of skill utilization and workload, on various indicators of worker strain. The results indicated that there was neither multiplicative nor adhtive effect ofjob control/skill utilization and workload on worker strain. Some of the relationships between job characteristics and worker strain were not stable over time. The cross-lagged correlation analysis showed that workload at Time 1 had some effect on one ofthe worker strain indxators at Time 2, but not the reverse.