Forty Hispanic junior high school girls divided into target and comparison subsamples were matched with target and comparison teacher subsamples to form student-teacher dyads. Dyad members simultaneously recorded their own and each other's behavior on a 41-item behavioral checklist. The data were analyzed to determine differences in behavior reported for the four subsamples, the effects of the various dyadic combinations upon behavior, and the extent of agreement between student and teacher raters. The behavior of the target girls was found to differ significantly in target and comparison teacher classrooms. While students indicated a difference in the behavior of teachers towards comparison versus target girls, teachers indicated no difference in their behavior towards the two groups.