This study reported methodology and results of a longitudinal evaluation of elementary students. With segregated classroom settings used as controls, students were placed in each of the following experimental classroom types: (1) those with size reduced and compensatory features involving primary-level students in an all-black school; (2) those where intermediate grade black children were transferred to several receiving schools with all white enrollment; and, (3) those which brought white pupils into inner city schools and provided for voluntary transfer of inner city pupils to outer city schools. Findings of the study concluded that students in segregated settings (control) achieved less than did pupils in compensatory or integrated settings. (Author/DM)