Students with emotional and behavioral problems are placed in a variety of classroom settings, presumably where their needs can be best served. Little attention, however, has been dedicated to monitoring their behavioral progress in these settings. This study analyzed data on student engagement, disruption, and destruction during 10 types of instructional activities. Data were collected at two times across one academic year in Pennsylvania and California in 90 classrooms, ranging from grades 1 to 12. Tests of mean comparisons (i.e., independent samples and paired t-tests) and the association between variables (i.e., bivariate correlation analysis) were conducted to address a number of concerns regarding the status and stability of classrooms for students with EBD. Independent paired sample t-tests showed a lack of significant improvement in behaviors across time, and the correlational data reflect a significant relationship between student engagement and classroom activity. Findings are discussed in light of limitations and directions for future research. In addition, practical implications are noted.