The present study evaluated the efficacy of a multicomponent, classroom-based intervention in reducing preschoolers' behavior problems. The Chicago School Readiness Project model was implemented in 35 Head Start classrooms using a clustered-randomized controlled trial design. Results indicate significant treatment effects (ds = 0.53-0.89) for teacher-reported and independent observations of children's internalizing and externalizing behavior problems. Moreover, there was some evidence for the moderating role of child gender, race/ethnic group membership, and exposure to poverty-related risk, with stronger effects of intervention for some groups of children than for others. Findings contribute to a growing area of research on poverty and preventive intervention in early childhood. (author abstract)