Reported in this paper are the findings of five meta-analyses assessing the relationships among children's interpersonal cognitive problem-solving (ICPS) skills, training, and behavioral adjustment. Five hypotheses were examined: (1) Adjusted children score higher on ICPS measures than do nonadjusted children; (2) Children trained in ICPS skills demonstrate a higher level of ICPS skills at posttesting than do controls; (3) Teachers' posttest behavior ratings are more positive for trained children than for controls; (4) Social behaviors observed at posttesting are more positive for trained children than for controls; and (5) A direct relationship between increased ICPS skills and improvement in behavioral adjustment can be substantiated. Findings indicated that effects exist for the first three hypotheses. Magnitude of difference for the fourth and fifth hypotheses was not large, and the small magnitude of effect for the fifth hypothesis casts doubt on the application of the analytical theory. To extend the analysis, boundary conditions were investigated, and studies involving only special populations were examined. It was concluded that while the meta-analytical model and interventions delineated by Shure and Spivack (1979, 1980, 1982) do show reliable, if not always large, effects across studies, continued replication and refinement are needed. (RH)