Abstract: One important function of close relationships is to regulate health behavior through social control. Social control affects health behavior through overt pressure (direct social control) and through feelings of responsibility for another that motivate healthy behavior (indirect social control). This project begins with a review and critique of the social control literature, organized around several social control models. Support for the models is mixed. Therefore, a meta-analysis of the social control literature was conducted. Results of the meta-analysis indicate that direct and indirect social control are positively associated with beneficial health behavior and with positive affect. Direct social control directly predicted negative affect, and indirect social control inversely predicted negative affect. Based on the literature review and the findings of the meta-analysis, a new theoretical model of social control is proposed. The model (a) differentiates direct and indirect social control, (b) bifurcates direct social control into communicative and physical direct social control, and (c) incorporates the message features of dominance, explicitness, and argument into communicative direct social control. Social control is predicted to influence affect, cognition, and health behavior. Relational interdependence was predicted to moderate the relationship between social control and health behavior. A primary data study was conducted to test the new social control model. Subjects (N = 101) were diabetics participating in diabetes education classes, who completed baseline, follow-up, and diary surveys. Though several of the predicted relationships were not supported, results showed that health behaviors could be improved by increasing the extent to which people consider the benefits of engaging in appropriate behaviors. Communicative direct social control, indirect social control, and argument positively predicted considering these benefits. A separate measure of health behavior was inversely influenced by negative affect, which was positively associated with dominance and negatively associated with indirect social control. Finally, the relationships between communicative direct social control and health behavior and between argument and health behavior were moderated by relational interdependence, such that the relationship was stronger at higher levels of interdependence. Implications for theory and practice are discussed, as are the strengths and weaknesses of the study and possible directions for future research.