This study tested a model of the relationship between social-cognitive variables and physical activity in a sample of 277 university students using a prospective design. Results of structural equation modeling indicated a good fit of the social-cognitive model to the data. Self-efficacy had the greatest significant total effect on physical activity, largely through its significant association with self-regulation, which directly predicted physical activity. Social support indirectly predicted physical activity through its significant association with self-efficacy. Outcome expectations had the smallest total effect on physical activity. Overall, the social-cognitive model explained 51 percent of the variance observed in physical activity.