Social and affective objectives are being increasingly accepted as integral parts of the school curriculum. Generalization effects produced by a training program in classroom participation skills were analyzed for two classrooms of seventh-grade boys with reading, writing, and disciplinary problems. Subject, situation, and behavior variables were combined to study eight different types of post-treatment generalization. The first classroom of students (N=14) received regular classroom instruction. The second classroom was composed of two experimental groups, i.e., 7 students who participated in the training program and 7 students who were exposed only to training program students. Teachers and trained observers rated students on 19 specific behaviors occurring in French and mathematics classes before treatment, and at 1- and 3-month intervals after treatment. Results indicated that training in classroom participation skills modified the behavioral repertoire of both training and exposure groups. Findings suggest that understanding and anticipating diverse types of generalization effects is necessary to measure the total impact of intervention programs. (Author/NRB)