This study proposes a model of associations between young children's social cognition and their social behavior with peers. In this model, two latent structures -children's representations of peer relationships and emotion regulation -- predict children's competent, prosocial, withdrawn, and aggressive behavior. Moreover, the model proposes that links between these two latent structures and children's social behavior are mediated by three discrete social-cognitive processes: encoding of social cues, hostile attributions, and social strategy generation. It was further hypothesized that the discrete social-cognitive processes would be associated with social behavior when children's receptive vocabulary was controlled, and that encoding of social cues, hostile attributions, and social strategy generation would make independent contributions to the prediction of social behavior. Subjects were 83 4- and 5-year-old children who completed multiple assessments that were developed or adapted to measure cognitive representations of relationships and each of the three discrete social-cognitive processes. Mothers and teachers rated children's emotion regulation, and teachers rated children's social behavior. Results indicated that, by and large, associations between the discrete social-cognitive processes and social behavior remained significant when verbal ability was controlled. However, there were fewer significant associations between discrete social-cognitive processing variables and children's social behavior. Moreover, there was no evidence to support the hypothesis that discrete social-cognitive variables would make unique contributions to the prediction of social behavior when other aspects of discrete social-cognitive processing were controlled. Measures of cognitive representations of relationships and emotion regulation were associated with measures of social behavior. Evidence supporting two of five hypothesized mediational paths was found for one of the two measures of cognitive representations. Evidence for an additional mediational path that was not hypothesized also was found. No evidence was found for mediational paths from the second measure of representations of relationships. No evidence supporting mediational models involving emotion regulation was obtained. Post-hoc analyses suggested that associations between hostile attributions and aggressive behavior was moderated by emotion regulation. Results suggest that pathways connecting children's emotion regulation, representations of peer relationships, discrete social cognitions, and social behavior are complex, specific, and interacting. Appended are: (1) Letter of Informed Consent; (2) Instructions for completing the peer affiliation assessment; (3) Feelings About Myself and Peers Puppet Interview; (4) Children=s behavior questionnaire; (5) Descriptions of the 10 videotaped vignettes; (6) Instructions for conducting and coding videotape-based interview; (7) Enactive and line-drawing based interview; (8) Teacher=s Checklist of Peer Relations; and (9) Preschool Socio-affective Profile - Short Form. (Contains 14 tables and 5 figures.) [Ph.D. Dissertation, Auburn University.]