The present study used data for 30 classes across 10 middle and high school teachers to evaluate the stability of class-level ratings on the Responsive Environmental Assessment for Classroom Teaching across time. Teachers collected data on 2 occasions and students' ratings (N = 806) were aggregated to the class-level. Classes were arranged into 2 groups on the basis of the latency between repeated administrations (M[subscript groups] = 8.1 and 15.1 weeks). Correlations between class ratings were slightly larger among classes in the first group (range = 0.73 to 0.87) relative to the second group (range = 0.62 to 0.75) which had a larger gap in time between data collections. Results provide evidence for the stability of students' perceptions of the instructional environment when those ratings are interpreted at the level of a class and support ongoing work to establish norms and benchmarks to guide interpretation.