The utility and validity of Bandura's self-efficacy construct was evaluated in a study involving 62 volunteer overweight women. Pretreatment, posttreatment, and follow-up assessments of self-efficacy and weight loss were obtained in the context of a behaviorally oriented weight-reduction program. It was predicted that (a) weight loss would be related to enhancement in personal efficacy, (b) posttreatment efficacy would be a significant and more accurate predictor of weight loss during follow-up intervals than posttreatment weight loss, and (c) self-efficacy would be related to attrition whereas weight loss would not. Results indicate a significant relationship between weight change and personal efficacy during the follow-up intervals. In accord with Bandura's theory, efficacy expectations predicted outcome during the 6-week and 6-month follow-up intervals, whereas posttreatment weight loss was unrelated to later outcome. In addition, completers had a substantially higher level of personal efficacy than dropouts over treatment and, to a lesser extent, over the follow-up period. Some procedural considerations regarding the assessment of personal efficacy are raised, and the relationship between the cognitively based constructs of goal choice and self-efficacy and task performance are discussed. Finally, given the nature and variability of weight-reduction efforts and the overall time span of the study, it is concluded that results attest to the validity and utility of the self-efficacy construct in predicting behavioral persistence and level of behavioral change.