OBJECTIVE To study the seroprevalence of human T-lymphotropic virus type I (HTLV-I) in a sexually active population and to determine sexual behavior risk factors for infection. METHODS Cross-sectional seroprevalence study using enzyme-linked immunosorbent assay (ELISA) and Western blot. Risk-factor data were gathered by administered questionnaire and chart review. METHODS Two urban, primary care clinics for persons with sexually transmitted diseases run by the Jamaican Ministry of Health. METHODS Of the 2050 consecutive patients presenting with new episodes of sexually transmitted disease, 1977 patients were eligible for analysis. RESULTS Overall HTLV-I seroprevalence was 5.7%; prevalence increased with age from 1.6% (age, 14 to 19 years) to 5.1% (age, 30 years and older) in men and from 5.3% (age, 14 to 19 years) to 14.1% (age, 30 years and older) in women. Compared with a reference cohort of food service employees, age-adjusted HTLV-I seroprevalence was increased in female patients with sexually transmitted disease (odds ratio = 1.83; CI, 1.41 to 2.83) but not in male patients with sexually transmitted disease. Independent risk factors for HTLV-I infection in women included having had more than ten lifetime sexual partners (odds ratio = 3.52, CI, 1.28 to 9.69) and a current diagnosis of syphilis (odds ratio = 2.12; CI, 1.12 to 3.99). In men, a history of penile sores or ulcers (odds ratio = 2.13; CI, 1.05 to 4.33) and a current diagnosis of syphilis (odds ratio = 3.56; CI, 1.24 to 10.22) were independent risk factors for HTLV-I infection. Of 1977 patients, 5 (0.3%) had antibodies to human immunodeficiency virus type 1 (HIV-1), including 2 with HTLV-I and HIV-1 coinfection. CONCLUSIONS We conclude that HTLV-I is transmitted from infected men to women during sexual intercourse. Our data are consistent with the lower efficiency of female-to-male sexual transmission of HTLV-I, but penile ulcers or concurrent syphilis may increase a man's risk of infection.