The purposes of the study were twofold: to determine whether educable mentally retarded (EMR) students are rejected during overt interactions with nonEMRs, and to determine whether EMRs who were integrated full-time in a nongraded school were perceived by their nonEMR peers to be similar to segregated EMRs or nonEMRs. Forty-two fourth, fifth and sixth graders were asked to select one of two children as a partner to help them win a prize at a bean toss game. Depending upon the treatment, the other two children were either: a segregated EMR and a nonEMR, an integrated EMR and a nonEMR, or segregated EMR and an integrated EMR. The results indicated that both integrated and segregated EMRs were chosen less often than nonEMRs, and that integrated and segregated EMRs were selected equally often. The findings were discussed in terms of the competence versus liking dimension. Also, it was suggested that future investigations might examine the effects of physical deviance on the EMR's social acceptability. (Author)