Objective: To assess the levels and correlates of potential exposure to and transmission of HIV in a contemporary, community-based probability sample of men who have sex with men (MSM).Methods SINCE THE MID-1990s, SUBSTANTIAL increases in the rates of bacterial sexually transmitted diseases (STDs) have been reported in men who have sex with men (MSM) in many industrialized countries. 1-6 Some data suggest that the incidence of human immunodeficiency virus (HIV) infection may also be rising in this population. 7,8 Recent reports indicate that risky sexual behavior in MSM-as reflected by numbers of sex partners, unprotected anal intercourse (UAI), and other measures-may have been increasing during this time as well. 3,4,9,10 Most investigators have attributed these trends directly or indirectly to improved treatment of HIV infection and the acquired immunodeficiency syndrome (AIDS) with highly active antiretroviral therapy (HAART). However, all recent data on sexual behavior in MSM have come from convenience samples, typically MSM with reported STDs, attendees of public STD clinics, persons receiving clinical or prevention services, or MSM contacted through outreach. Except for general population surveys that included relatively few MSM and limited assessment of MSM sexual behavior, 11,12 the only reported population-based surveys of sexual behavior in MSM were performed before the development and widespread use of HAART. Therefore, we undertook this study to estimate the frequency and correlates of sexual behaviors conducive to acquisition or transmission of HIV in a probability sample of MSM.