A primary aim of the Chicago School Readiness Project was to improve teachers' emotionally supportive classroom practices in Head Start-funded preschool settings. Using a clustered randomized controlled trial (RCT) design, the Chicago School Readiness Project randomly assigned a treatment versus control condition to 18 Head Start sites, which included 35 classrooms led by 94 teachers who served 602 children. Teachers in the treatment condition were invited to participate in behavior management training and their classrooms were visited weekly by mental health consultants who ''coached'' teachers as they implemented behavior management strategies. Estimation of hierarchical linear models revealed that the multi-component intervention provided statistically significant benefits: Head Start classrooms randomized to the treatment condition were found to have statistically significantly higher levels of positive classroom climate, teacher sensitivity, and behavior management than were classrooms in the control condition (with effect sizes ranging from d=0.52 to 0.89). Discussion of these findings reflects on policy implications and future research.