In this study, we assessed the stability over time of child, parent, and teacher reports of school climate using one widely used assessment that provides forms for child, parents, and teachers: the School Development Program-School Climate Surveys. To assess whether measures of school climate from multiple informants differentially predict child outcomes, we examined concurrent and predictive associations of each of these sources of data with children’s peer victimization (child report) and internalizing and externalizing problems (parent and teacher reports). Participants included 1237 elementary school children and their parents (n = 1082), and teachers (n = 964). Data were collected in the fall and spring across two academic years (4 waves of data). Confirmatory factor analyses showed that the sub-dimensions for each informant version of the scales adequately fit a single factor model that was invariant over time and for child sex. Differences in informant source were important in predicting child outcomes. Implications for targets for improving school climate are discussed.