To investigate the assumptions that interpersonal problem solving skills can be systematically taught to young children and that such skills mediate behavioral adjustment, 37 black, economically disadvantaged preschool age children participated in a year-long intervention project and a 6-month follow-up. Specifically, the study evaluated the long-term effectiveness of the Spivack and Shure training program in bringing about changes in cognitive interpersonal problem solving and behavior. The intervention program consisted of a series of brief, daily "games" involving small groups of children. Games initially focused on language skills related to alternative, consequential, and causal thinking, then explored individual's likes and dislikes, culminating in the application of skills and concepts in the contexts of real life situations. After training with the games, subjects were seen three times for individual testing by research assistants who administered the Preschool Interpersonal Problem Solving Test, the What Happens Next? Game, and the Caldwell Preschool Inventory. Additionally, each child's behavior was observed on three separate occasions by trained observers using the Sharp Behavior Checklist. The Hahnemann Preschool Behavior Rating Scale was completed for each child by the classroom teacher after she had known the child for 3 months. Findings indicate that behavior change in young children cannot be mediated through a strictly cognitive intervention and that behavior change requires an integration of behavioral and cognitive techniques. (Author/RH)