The study evaluated the reliability and validity of SCAN, a measurement technique for assessing class involvement in the learning process (classroom climate for learning). Concurrent validity was investigated by comparing SCAN ratings to a point-sampling of on-task performance behaviors. The Engaged Time Observation Instrument (ETOI) was used to record both individual student behaviors and SCAN data. Reliability of the SCAN was determined through comparison of the recordings of two independent observers. A total of 185 observations (mean length of time 41 minutes) was made of 94 regular class students (grades 2-6) from six elementary schools in two New York City school districts. Students included 39 without handicaps and 55 with mild learning problems who were receiving resource room services. Correlations of SCAN values with actual on-task behavior of the entire sample of individual students indicated a moderate relationship (.62) between measures. Expected differences in classroom atmosphere for learning between the two school districts were also reflected in the absolute values of the SCAN means, indicating that the technique can discriminate between environmental setting variables. It is concluded that SCAN is a useful technique for conducting observational assessments of children's behavior in relation to the quality of the classroom environment for learning. (JW)