A meta analysis of 93 research reports on school desegregation and black achievement was conducted to determine (1) why the study results differed as to apparent desegregation effects; and (2) whether some desegregation plans produced greater achievement gains than others. An analysis indicated that inconsistencies in research findings were mainly due to two methodological flaws: the inability of most desegregation studies to focus on the earliest levels of schooling, where desegegation effects are most significant; and the absence of adequate control groups in most studies. Examination of reported desegregation effects showed that (1) desegregation produced greater gains for IQ scores than for achievement test scores; (2) achievement gains in successful desegregated schools were greater in reading and language arts for elementary school pupils and in science and social studies for high school students; and (3) the most successful desegregation plans were those that resulted in schools with a majority and more than a token number of blacks. Implications for theory, public policy, and educational research were drawn from the analysis. (Author/MJL)