The purpose of this study was to determine if sexual orientation can be correctly identified under controlled conditions. A series of 24 brief videotaped interviews with homosexual and heterosexual men and women were presented to a sample of 143 subject raters divided into four sexual preference and gender groups. None of the groups were able to exceed levels of correct detection. Approximately 20% of the total subject pool did exceed chance levels. There were significantly more women than men in this sub-sample and homosexual women were represented disproportionately. Although there were some differences in the types of behavioral cues used by the different sexual preference groups to make their judgments, with the possible exception of homosexual women, these cues were unrelated to accurate identification of sexual orientation. The relatively better performance of female raters is discussed in terms of differences in the socialization process of men and women.