Purpose: The purpose of this presentation is to describe the results of a three-year study of a prevention program intended to promote social competence and prevent aggressive behavior in third-grade children. Based on social information processing (SIP) and emotion regulation theoretical perspectives, a skills-training program was developed and delivered to 3rd-grade students (51% male) in two schools comprised of African American (20%), Latino (41%), non-Latino White (34%), and other (5%) children. Entering 3rd grade in 2000, 182 children received a routine health curriculum. Entering 3rd grade in 2001, 173 children received the Making Choices (MC) program delivered by intervention specialists who visited schools once per week. Entering 3rd grade in 2002, 198 children received the MC program plus an intervention designed to involve teachers and parents in MC-related activities. The teachers in 2002-3 were given supplemental activities to reinforce and infuse MC training in routine classroom activities. In addition, parents were invited to a series of five "family nights," school-based evening programs designed to promote parent-school involvement, and they received a monthly newsletter that included home-based MC exercises. Methods: Teacher ratings of student behaviors and child assessments of SIP skills were collected in each of the three years during which the study was conducted. Behavioral outcomes measures include: cognitive concentration, social competence, social contact, social aggression, authority acceptance, and overt aggression. SIP-related skill acquisition was assessed using posttest analogue methods. Measures included: encoding, hostile attribution, goal formulation, and response decision-making. Two- and three-level hierarchical linear models were used to estimate treatment effects. Level-1 regresses student posttest scores from each measure on covariates - pretest score, race/ethnicity, gender, and risk exposure. Level-2 regresses the student-level intercept and, where fit improved, slope estimators on classroom (j=29) indicators for school and intervention. For social and overt aggression, three-level models (student, classroom, and teacher) were estimated. Results: No significant differences across cohorts were observed at pretest. Controlling for covariates and testing for cross-level interactions, children in classrooms receiving the interventions had fewer behavioral problems at posttest when compared to the comparison cohort. However, findings varied by the type of intervention, i.e., MC alone versus the augmented MC. In contrast to the comparison group, MC alone had significant effects on children's social contact, social aggression, and overt aggression. The augmented MC had significant effects on cognitive concentration, social competence, social aggression, and overt aggression. SIP skills mediated the effect of both MC conditions on behavioral outcomes. Implications for Practice: Findings suggest that strengthening children's skills in regulating emotions and processing social information reduces conduct problems associated with social aggression and physical aggression. A broader pattern of effects can be observed when skills training content is infused into regular classroom activities and when parents can be involved in training-related content.