The object of this study was to determine the extent to which disadvantaged kindergarten pupils would benefit from association with advantaged pupils and teachers in the improvement of basic skills and self-concept. The sample for 1967-68 was composed of 20 disadvantaged rural Negro and Caucasion 5-year-old children randomly selected. The sample for 1968-69 was reduced to 16 children. In each instance these children were integrated with approximately 30 foreign and Caucasian middle class children. In 1967-68 only the disadvantaged children were administered pre- and posttests approximately 7 months apart. Comparison of results showed significant gains in total performance, language skills, and self-concept, while IQ scores yielded nonsignificant results. In 1968-69, evaluation focused on comparisons of experimental (deprived) and control (underprived) means for pretests, posttests, and mean gains from pre- to post for each group. The results must be viewed with caution, for although scores did not show that the disadvantaged gained more, they did gain as much as the advantaged. More statistically significant divergence in favor of the control group was found between pretest results than posttest results. This reduction of difference could have been due to a combination of socialization and curriculum. (JF)