Recent studies indicate that men are more homophobic than women and that a negative bias among men is particularly evident toward homosexual men (Augero, Bloch, & Byrne, 1984; D'Augelli & Rose, 1990). Self-report measures have frequently been used to indicate homophobia and are susceptible to distortion by respondents. Therefore, the present study used the "wrong number technique" (Gaertner & Bickman, 1971) to obtain a nonreactive measure of homophobia. Forty male and forty female subjects received an apparently wrong number telephone call from a male caller who portrayed himself as either a heterosexual or a homosexual. Indicating that his car had broken down and that he was out of change at a pay phone, the caller requested help by asking the subject to call his girlfriend/boyfriend for him. It was expected that more help would be given to perceived male heterosexuals than to perceived homosexuals, particularly among male helpers. Results indicated both males and females were significantly more likely to help male callers perceived to be heterosexual (p <.01), but helping by males and females did not differ in the two sexual orientation conditions. Results are discussed in terms of current information about homophobia in society.