This study compared the ability of Bandura's social cognitive theory and Fish-bein and Ajzen's theory of reasoned action to predict exercise behavior. The theories' constructs were assessed and then the exercise behaviors of 328 individuals were recorded for the following 7 weeks. A path analysis indicated that the theory of reasoned action model fit the data, but explained only 5 % of the exercise behavior variance. Two social cognitive theory variables, self-efficacy and self-evaluated dissatisfaction, significantly predicted exercise behavior. Also, a multiplicative function of self-evaluated dissatisfaction and outcome expectations increased the amount of predicted exercise behavior variance to 16%. Thus, individuals who were confident they could adhere to an exercise program and were satisfied with their standing on probable outcomes from participation (e.g., present body weight) exercised more days per week. A commonality analysis indicated that the theory of reasoned action did not account for any unique variance in exercise behavior over the social cognitive theory constructs. In sum, social cognitive theory was more effective than the theory of reasoned action in predicting exercise behavior.