The effectiveness of a cognitive-behavioral interpersonal problem solving (IPS) training program was evaluated with 35 poorly accepted second, third, and fourth graders. Group 1 received instruction in IPS and included only Ss low in peer acceptance; group 2 consisted of Ss low in peer acceptance who participated in IPS with same sex and grade Ss high in peer acceptance; and group 3 were contact control Ss low in peer acceptance who did not receive IPS but participated in games and activities. Dependent measures included the Social Problem Solving Test, Self-Evaluation Loneliness Scale, Classroom Inventory, and Teacher Rating of Peer Acceptance. IPS training focused on five major aspects; (1) establishing rapport, (2) identifying and using feelings, (3) understanding solutions in the long-run and short-run; (4) using four steps to solve problems, and (5) applying problem solving to group situations. Preliminary results pointed to significant differences in favor of the low social skills high/social skills (LS/HS) group when compared with the contact control group. There were no significant differences on peer ratings of acceptance between LS and LS/HS Ss. A reference list is appended. (CL)