Hypotheses put forth by McCarthy and Yancey are revised to suggest that the relationship between the self-evaluations of whites and blacks vary according to the trait. For some traits, such as performance of family roles, the self-ratings of blacks are at least equal to those of whites. For others traits, such as trustworthiness, whites rate themselves higher. It is suggested that these patterns are due to variations in (1) the significant others used, (2) the likelihood of using subcultural standards, and (3) the "availability" of a system-blame explanation.