Past research on school-level factors that predict children's development has focused largely on associations between a limited number of characteristics, such as school size and school resources, and children's academic achievement. Few studies take a more comprehensive look at the measurement of school climate or examine its relationship to children's social-emotional competence. Studies that aim to link features of schools with student outcomes typically necessitate a multilevel approach because students are nested in schools. Unlike many other studies of early elementary school, this study includes reliable measures of children's social-emotional competence. In the current paper, the authors capitalize on these strengths and the measurement capabilities of structural equation modeling to develop a new, multidimensional model of school climate, which they use to predict low-income children's social-emotional development during the transition to kindergarten. They use follow-up data from a cluster-randomized controlled trial of a classroom-based intervention in Head Start classrooms. The purposes of this study were to: (1) Identify a multidimensional model of school climate and (2) use it to predict low-income children's social-emotional outcomes during the transition to kindergarten. The research setting consisted of kindergarten classrooms located in the Chicago public schools (CPS). Descriptive statistics revealed substantial variation in children's social-emotional functioning and school characteristics at kindergarten. Preliminary results from 2-level unconditional hierarchical linear models (HLM) models suggest that a small but significant portion of the variance in children's social-emotional functioning was attributable to between-school differences (ICCs ranged from 0.09 to 0.25). Additional 2-level HLM analyses in which children's conflict with the teacher, closeness with the teacher, and social competence in kindergarten were predicted from a set of school characteristics and child-level controls indicated that a large school size was associated with a small but significant increase in teacher-child conflict (B = 0.003, p less than 0.05) and a small but significant decrease in children's social competence (B = -0.004, p less than 0.01) between preschool and kindergarten. In contrast, a large percentage of children with an Individualized Education Plan (IEP) was associated with a marginally significant decrease in teacher-child conflict and a marginally significant increase in children's social competence. The inclusion of controls for children's social-emotional functioning in Head Start makes these models rigorous and conservatively specified, allowing for greater precision in the authors' estimates. (Contains 2 tables.)