The research reported in this paper has three major purposes: to assess the development of racial attitudes in urban children, to evaluate the effects of race of examiner on children's expressed attitudes, and to examine some of the perceptual correlates of racial attitudes. It was predicted that older children would perceive faces of another race as less distinctive from one another than would younger children. Two interrelated studies were conducted. The first was a testing program which assessed the racial attitudes of second, fourth, and sixth grade children. The variables were, in addition to chronological age: race of subject, race of examiner, and type of testing instrument--direct questionnaire, Social Distance Scale, and projective. The second study selected a subsample of these children, who were instructed to judge the similarity of schematic drawings of facial pairs which varied systematically along a number of stimulus dimensions. An apparent decline in prejudice with age was obtained on a direct questionnaire and Social Distance Scale, but was not evident on a more indirect measure. This suggests that the effects of social desirability must be considered, even in measuring the attitudes of young children. [Not available in hard copy due to marginal legibility of original document.] (Author/JM)