The impact of a school desegregation plan on the social class structure of elementary classrooms (Grades 1-6) was studied within a large midwestern school system. Sociograms administered on a pre/post basis, over a two year period, to all classrooms with at least four minority students per room served as the data base. Initial analyses reveal separation on the basis of sex to be highly significant across all grade levels for all years. A stricter test of choice patterns was conducted by subsetting the data to those classrooms containing at least four minority students of the same sex. Results of this test did not confirm a relationship between segregative choice patterns and the percentage of minority students in a classroom; i.e., the higher the percentage of minority students in a classroom the more likely the segregative choices being made by students, was not evident. (Author)