Two issues which are addressed in children's cross-race relationships are assessment and intervention. Results from two different types of sociometric studies are reviewed: those using measures of cross-race "best-friendship" and those using measures of "acceptance." Comparisons suggest that using a criterion of best-friendship rather than acceptance provides an overly negative picture of children's racial attitudes. Data are presented from a recent longitudinal study of cross-race acceptance. The 38 black and 116 white children who participated had experienced desegregated education since kindergarten. They were tested when they were in third grade and again when they were in sixth grade. Children were asked to indicate on a five-point scale how much they liked to play with each of their classmates. Both black and white children showed some preference for their own race in third grade. By sixth grade own-race preference increased, especially among black children. The degree of racial bias was small, however, and the results provide a more positive picture of cross-race relationships then that obtained in previous research. Programs designed to improve children's cross-race relationships are reviewed and suggestions are made for further research on the social-cognitive bases of children's cross-race liking. (Author)