This study investigates whether the racial composition of classes attended by black students has differing effects on their academic performance under different conditions. Specific variables examined are friendship patterns with whites, characteristics of white peers, and characteristics of the black students themselves. Also examined is the relationship between racial composition of high school classes and academic performance for those students in different curriculum programs and for those who attended elementary school classes with varying racial composition. The authors report that the results show little support for the hypothesis that interracial contact will benefit blacks academically when they are placed with white students of higher academic attainment. The association between interracial contact and academic outcomes for blacks is reported not to differ appreciably for those with differing amounts of friendship with whites, for those who attended grade schools with a different racial composition, or for boys as compared to girls. However, it is reported that placement in largely white classes appears to benefit those black students whose black peers are not academically oriented, and those whose white peers' parents and whose own parents had little education. The authors attribute this to the possibility of different expectations and demands from teachers. (Author/WP)