The convergent and discriminant validity of a wide range of cognitive and social competence measures were investigated with 400 students (7-11) stratified by race (white and black) and educational placement (regular and special education for mildly handicapped students). The majority of special education students were classified as learning disabled. As expected, the special education Ss scored substantially below regular education Ss on all academic competence indices (including the Wechsler Intelligence Scale for Children Revised and the Peabody Individual Achievement Test). The magnitude and pattern of differences among special and regular education students were quite similar for black and white Ss. Special education Ss obtained substantially lower scores on all social competence indices (including classroom sociometrics and the Adaptive Behavior Inventory for Children). The magnitude of differences among regular and special education students suggested relatively little overlap of distributions for white and black students. Discriminant function analysis suggested relatively strong convergent and discriminant validity using the criterion of handicapped child vs. regular education placement. A multitrait-multimethod analysis of the academic and social competence measures yielded less positive results, suggesting relatively high method variance and relatively low trait variance for the social competence measures. Among implications noted are that social competence measures, particularly with minority students, are likely to be valuable adjuncts to traditional assessment practices. (Author/CL)