Two-hundred-and-sixty undergraduates at a two large eastern universities (who self identified as black or white) completed a confidential anonymous twenty-two-item questionnaire designed to assess perceptions of women and men. In general, while both races tended to view women and men of their own and the other race positively, there was a pronounced tendency to view women and men of one's own race more positively and members of the other race more negatively. Social network analysis and the contact hypothesis were used to explain the findings--the races tend to self-select into their own groups and such selection is not conducive to understanding other races and questioning stereotypes. Limitations and implications of the data are suggested.