The present study examines data collected during the second year of a three-year longitudinal cluster randomized controlled trial, the Responsive Classroom Efficacy Study (RCES). In the context of and RCT, the research questions address naturally occurring variability in the independent variables of interest (i.e., teachers' (fidelity of implementation) FOI to RC practices) and their relation to students' mathematics achievement. Findings highlight the differences between traditional measure-based composites of FOI and contemporary practice-based composites, as well as the relation between teachers' use of specific RC practices and students' mathematics achievement, but do not permit causal inferences regarding these associations. (Contains 3 figures and 5 tables.)