This study compared characteristics of patients who had herpes simplex virus (HSV) type 1 with characteristics of patients who had HSV-2, by use of data from a cross-sectional analysis. Data were collected in an urban sexually transmitted diseases clinic from patients who had positive genital HSV cultures. Overall, 17.1% (95% confidence interval [CI], 14.9%-19.3%) of 1145 genital HSV isolates obtained during 1993-1997 were HSV-1. The proportion of HSV-1 among initial genital herpes infections was higher among men who had sex with men (46.9%) than among women (21.4%) and was lowest among heterosexual men (14.6%). White race (odds ratio [OR], 3.7; 95% CI, 2.3-5.9) and receptive oral sex in the preceding 2 months (OR, 2.8; 95% CI, 1.9-4.3) significantly increased the odds that initial infections were HSV-1 rather than HSV-2. Genital HSV-1 may often be acquired through contact with a partner's mouth. These data suggest that seroprevalence studies based solely on HSV-2 type-specific assays underestimate overall prevalence of genital HSV infection.