This paper analyzes 19 studies that were considered by the National Institute of Education's panel on the effects of school desegregation in order to determine desegregation's effect on black achievement, black self-esteem, and racial relations. It is stated that, overall, the results show that reading achievement improves somewhat as a result of desegregation but math achievement does not. The author notes that there is a basic problem in evaluating desegregation programs in that there is great diversity among programs, which accounts for the diverse results obtained in different studies. It is suggested that desegregation has not increased black self-esteem, and in some cases has decreased it; tentative conclusions suggest that racial relations have not been improved by desegregation. According to the author, research, and particularly long-term research, is still needed on the effects of desegregation on students and communities, as well as studies that examine under what conditions desegregation does work. (CMG)