The purpose of this study was to design an instrument to measure interpersonal racial attitudes among primary children in segregated and nonsegregated school settings. Subjects were 129 first and second graders enrolled in a lower-middle class all white school, an all Negro slum school, and an integrated lower-middle class school. All received test booklets with decals of a target figure (Negro teacher, white teacher, peer, or school) on each page, and a package of stickers on which self-figures were printed. They were directed to paste the self-figures in whatever position and at whatever distance from the target figure they chose. White subjects placed the self figures farther away from Negro targets than from white targets. Negro subjects placed themselves equally close to white and Negro targets. Integrated Negroes placed themselves significantly closer to white-children targets than did segregated Negro subjects. Figure placement indicated integration accelerated a tendency for white subjects to move closer to Negro children targets as grade increased. Regardless of race, integrated subjects placed themselves closer to school than did segregated subjects. Because this study was preliminary and small, many interpretations exist and cannot be resolved for some findings. Further research is needed to compare the validity of alternative interpretations. (DO)