Students often have differing perceptions of their school's climate. Although these subjective perceptions can be meaningful outcome predictors, discrepancies create challenges for those seeking to globally characterize or intervene to improve the climate of a school. Trained, outside observers can provide insights on perceptible and ostensibly malleable aspects of the school; however, the extent to which these outside observations help us to understand differences in students' report of school climate has not been examined. To study this, we examined convergence and divergence between independent observers' assessments and students' shared perceptions of school climate at the classroom- and school-levels. Data come from the Maryland Safe and Supportive Schools Initiative (MDS3), which included 20,647 students and observations of 303 teachers in 50 high schools. Students responded to survey items regarding safety, engagement, and environment; independent observers assessed teachers' classroom practices and the school physical environment. A three-level model partitioned variance to the individual, classroom, and school levels. Most of the variance in students' climate perceptions was between students, with classroom and school intraclass correlation coefficients (ICCs) ranging from 0.01 to 0.20; however, observations explained large percentages of between-school variation in climate (58% to 91%). Findings suggest the potential utility of outside observations for explaining school-level variation in school climate to inform decision-making and future research. [This is the in press version of an article published in "American Journal of Community Psychology" 2019.]