We examined the convergent validity of observer ratings of teachers' use of evidence-based instructional and behavior management practices using the Classroom Strategies Assessment System (CSAS; Reddy & Dudek, 2014) with student academic engagement as measured by the Cooperative Learning Observational Code for Kids (CLOCK; Volpe & DiPerna, 2010). This study was conducted with a sample of 107 teachers and 2,000 students in 11 urban elementary schools serving students in a community with high concentrations of poverty. Correlational and multiple linear regression analyses examined whether CSAS instructional and behavior management scores predicted classwide academic engagement. Results provide initial evidence of the relationship between observer ratings of teacher practices and student engagement. Specifically, CSAS scores indicating higher quality instruction and behavior management were associated with higher student engagement. Multiple linear regression revealed that instructional practice scores (as measured by the CSAS) uniquely predicted student academic engagement (i.e., attention and participation during instruction) while behavior management practice scores did not. Implications for future research and practice are discussed.