Research is described which explores the hypothesis that different classroom situations may be optimal for different individuals. The approach used cluster analysis to identify student and classroom "types" whose interactions were then examined in an analysis of variance framework. About 1,300 fourth graders from 50 classrooms were involved in the study which incorporated classroom observations, teacher comments, student questionnaires, and achievement tests. Cluster analysis resulted in the identification of six classroom types ranging from permissive, warm, individualized atmospheres to highly controlled, cold, nonindividualized atmospheres. Three student types were low achievers lacking self-direction and confidence, highly motivated self-confident achievers, and strongly autonomous self-directed moderate achievers. A summary of major trends shows that low achieving boys did best in warm classes with moderate control, highly motivated boys did best in controlled classes which allow for student initiative, and autonomous boys did best in permissive classes which allow for student initiative. Low achieving girls did best in warm and individualized classes; high achieving girls did best in classes combining warmth, control, and orderliness; and autonomous girls did best in warm classes emphasizing student expressiveness. Major sex differences are enumerated, also. (Author/AV)