The levels of performance on tests of black versus white students in 15 North Carolina public universities were examined; the validities of test scores and high school grades for predicting performance in college were also studied. Experimental variables obtained for each student included, as predictors, the SAT-V and SAT-M scores, and the standing of the student in his high school class of graduation. Criterion consisted of the student's first term grade average in college. For each of the four institutional sub-groups separately at each of the 15 institutions and then at each institution for all-black vs. all-white, all male vs. all female, and for all students combined, the zero-order correlations among the predictor and criterion variables were computed. Findings include: (1) Black students attending traditionally black universities have lower test scores than do black students attending traditionally white universities; (2) Black students at white institutions have lower scores than their white counterparts at that institution, but, in some instances, higher scores than white students at some other white institutions in North Carolina; (3) Black students at traditionally white institutions in North Carolina have higher standings in their high school class of graduation than do their white counterparts at that institution; and (4) the small number of black students found in traditionally white institutions in North Carolina prevent any conclusive interpretation of the data. (Author/CK)