The present study evaluated a brief intervention targeting reduction of heavy drinking in college females. Within the context of this brief intervention outcome study Bandura’s (1986) model of self-regulation was tested. Seventy-six heavy drinking college females (minimum of four drinking occasions during past month with Blood Alcohol Concentration estimated at .08% or greater) participated in the intervention. Subjects participated in brief individual assessment sessions and all subjects were provided with drinking reduction information and strategies. Subjects were randomly assigned to one of three goal-setting conditions (no goal; proximal drinking reduction goal, distal drinking reduction goal). Within each of these three goal conditions subjects were randomly assigned to receive feedback on their drinking behavior or to not receive drinking behavior feedback. Subjects were reassessed one month and two months later on measures of drinking behavior and self-regulation variables (commitment to not drinking heavily, efficacy for not drinking heavily and discrepancy/discomfort relative to drinking heavily). Contrary to hypotheses, goal-setting, the provision of feedback, or the combination of goal-setting and feedback was not superior to assessment and information in the reduction of heavy drinking. However, all conditions revealed a significant decrease in drinking across time. Although the self-regulation variables of efficacy and commitment related negatively to future drinking behavior in univariate correlational analyses, the interaction of the self-regulation variables (efficacy, commitment and discrepancy) failed to add to the prediction of future drinking beyond that accounted for by current drinking and the main effects of the self-regulation variables. The theoretically derived hypotheses were not supported by the present study. Procedural and theoretical limitations of self-regulation relative to reduction of heavy drinking in college females, as well as the difficulties involved in changing college student drinking given strong contextual influences are discussed. / Ph. D.