Based on a sample of nearly 8,000 college students, this study was undertaken to determine the prevalence of non-heterosexuality using four different measures of sexual orientation: a self-identity measure, an attraction measure, a fantasy measure, and an experiential measure. Over 97% of both males and females labeled themselves as heterosexual, with the proportion of homosexuals and bisexuals combined constituting nearly 3% of the male sample and about 2% of the female sample. Roughly 80-85% of both sexes would be classified as exclusively heterosexual, with about 10% of both sexes having at least half of their sexual fantasies involving same-sex partners. In terms of sexual experiences, about 5% of non-virgin males and 0.5% of non-virgin females reported that all of their experiences had involved same-sex partners. After examining each of the sexual orientation measures separately, we compared them to one another. These comparisons revealed an unsettling number of apparent contradictions in the responses given by individual participants (e.g., participants who stated that they were homosexual but that all of their fantasies involved members of the opposite sex). By eliminating participants who provided these apparent contradictory responses, a post-hoc refined self-identified measure of sexual orientation was created. This refinement had essentially no effect on our estimates of the proportion of college students who were identified as heterosexual or homosexual, thus suggesting that the inconsistencies were largely due to random response error. Boosting our confidence in the post-hoc refined measure, we found that it exhibited slightly stronger links with several established childhood correlates of sexual orientation (such as the enjoyment of collecting dolls and playing dress-up) than was true for the original self-identified sexual orientation.