Despite their widespread application, many of Rapaport, Gill, and Schafer's (1968) hypotheses regarding the Wechsler intelligence tests have not achieved empirical support. To test the assumption that the Comprehension and Picture Arrangement subtests are socially sensitive components of the Wechsler scales, individual subtest scores of the Wechsler Adult Intelligence Scale and Wechsler Intelligence Scale for Children were correlated with clinician ratings of social competence and hostility. In these analyses, normal subjects and subjects at risk for psychopathology were examined during both childhood and adolescence. Although some support for the hypothesized relationship regarding the Comprehension subtest was found for normal subjects in childhood, overall, the results did not support the assumption that the Comprehension or Picture Arrangement subtests are generally sensitive to social functioning. Implications of differences in patterns between at-risk and normal groups across ages are discussed.