This paper describes a study of the classroom behavior patterns of 79 reflective and impulsive elementary school children in each of three contextual settings: individual setting (independent, unsupervised seat work); large group setting (15 or more students instructed by the teacher or another adult); and small group setting (similar to large group, but with 10 or fewer students). Groups of 9- and 11-year-olds participated in the study. Fifteen behavioral observation categories were used. Eleven of these were discrete (only one category could be scored during any 5-second observation period), and four were non-discrete (capable of being scored with another category during an observation period). Examples of discrete categories were: independent work, distraction, and task-oriented discussion. Non-discrete categories included self-verbalization and out-of-seat for a non-task related purpose. Two 5-minute blocks of data were collected for each child in each of the three contextual settings. Results lend little support to the hypothesis that reflective and impulsive elementary school children display characteristically different patterns of task-oriented or social behavior in the classroom. They also cast doubt on the validity of the Matching Familiar Figures test as an index of reflective and impulsive classroom behavior. Hence, direct observational measures rather than MFF classifications are recommended for identifying impulsive children for classroom intervention. (CM)