Prior research demonstrates the importance of school climate in shaping student behavior but tells us less about which aspects of school climate matter. In this paper we consider how distinct elements of school climate relate to skipping school and cheating on tests. Using survey and administrative data from several statewide Delaware sources, we perform a series of random-intercept logistic regression models. We find that students in schools perceived to have a climate with high levels of structure and support are less likely to report cheating on tests. Yet we do not find a robust relationship between most climate measures and skipping school. School climate relates strongly to in-school deviant behavior but much less to school-related deviant behavior occurring outside of schools. By specifying what measures of climate do and do not relate to problematic student behaviors, our results sharpen our understandings of how school climate shapes student behaviors.