Research has demonstrated links between children's poor peer relations and various forms of child and adult psychosocial maladjustment. Social skills training programs have been developed to increase children's social competence and reduce the risk for later problems. The Sheidow Park Social Problem Solving Program is a curriculum based cognitive social skills training program, designed for Australian primary school children. The present research evaluated the effects of this program on a variety of dimensions of children's social competence. Subjects were Reception/Year 1 children in two classes of a South Australian suburban primary school. The teacher of one class implemented the social skills program, while the other class experienced no formal social skills intervention. The results indicated that the various measures of social competence employed were relatively independent of one another, supporting the need for a comprehensive range of measures in social skills training research. The Sheidow Park program demonstrated a significant effect on children's sense of social self-competence and the degree to which they perceived a variety of challenging social situations as difficult to deal with. However, the program had no effect on teacher and peer ratings of children's social competence or on children's satisfaction with their wider social network. The findings are explained within the context of attribution and cognitive dissonance theories, and the strengths and limitations of both the Sheidow Park program and the present research are discussed. Suggestions for future research and modifications to the program are made.