BACKGROUND: The causative factors of exercise-induced mood change remain unclear. The objective of this study was to clarify the relationship of exercise with change in vigor (energy level). METHOD:: Obese women in an exercise support treatment (n = 122) or control condition (n = 70) were assessed on vigor, exercise barriers self-efficacy, and physical self-concept at baseline and at month 6. Exercise session attendance was also assessed in the treatment group. The relationship between exercise attendance and vigor change was tested. Multiple regression analyses also tested whether change in vigor was predicted by exercise barriers self-efficacy and physical self-concept scores, both at baseline and change over 6 months. RESULTS: Vigor, exercise barriers self-efficacy, and physical self-concept were significantly improved in the treatment group only. Exercise session attendance was not significantly related to change in vigor. Multiple regression analyses indicated that exercise barriers self-efficacy and physical self-concept scores, both at baseline and change from baseline to month 6, explained a significant portion of the variance in changes in vigor (Radj = 0.13 and 0.21, respectively). In post hoc analyses, entry of depressed mood (indicated by a cut point on depression score) into the regression equations increased the variance accounted for to Radj = 0.15 and 0.25, respectively. CONCLUSIONS:: Findings supported behavioral explanations of the exercise-mood change relationship emanating from social cognitive and self-efficacy theory. The use of manageable amounts of exercise early in the weight loss intervention process was suggested. © 2010 Lippincott Williams & Wilkins.