This paper explores the comparative utility of institutional and personal qualities in explaining cognitive outcomes in self-esteem, racial attitudes, and political and social involvement in the context of desegregated schools. Three kinds of schools in a Mississippi county were studied: public desegregated, public segregated black, and private segregated white. A range of behaviors and attitudes about self and one's role in the school, community, and political systems was measured for 1,240 students in grades 6, 9, and 12, forming the dependent variables. The independent variables were hypothesized to be both institutional (type of school) and personal (race, status, maturity, sex, and interest in pursuing education). The type of institution was not the best predictor of cognitions and behaviors, nor was race or status. The strongest influence was the presence of an interest in schooling among older students, especially females, a group that is active, liberal with regards to race and gender, efficacious politically, and possessing a high self-esteem. Both races in desegregated schools had more racial contact and more positive racial attitudes than those in segregated schools, but the type of institution did little to predict self-esteem and political and social involvement compared to the much stronger influence of students' personal qualities. Seven tables and 22 references are included. (CJS)