Human T-cell lymphotropic virus type I (HTLV-I) has been etiologically associated with a neurologic syndrome called HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP) as well as with adult T-cell leukemia/lymphoma. The authors sought to quantify the risk in Jamaica of HAM/TSP associated with HTLV-I infection and cofactors associated with this disease among infected individuals. Between 1988 and 1989, prevalent and incident HAM/TSP patients and controls with other neurologic diseases were enrolled in a retrospective study. A second control group was composed of HTLV-I-seropositive, asymptomatic carriers in Jamaica, ascertained in a separate study conducted in 1988. Although HTLV-I seropositivity was not a component of the case definition for HAM/TSP, all 43 HAM/TSP patients were HTLV-I seropositive compared with two (4.0%) of the controls with other neurologic diseases. Given HTLV-I seropositivity, one cofactor associated with the risk of HAM/TSP was young age at initial heterosexual confidence interval 1.29-12.46 for individuals aged < or = 15; odds ratio = 4.26, 95% confidence interval 1.41-12.90 for individuals aged 16-17 years at initial intercourse). Among individuals who reported this early age at initial sexual intercourse, an increased risk of HAM/TSP was associated with having reported more than five lifetime sexual partners (odds ratio = 2.88, 95% confidence interval 0.90-8.70). Neither an early age at initial sexual intercourse or the number of lifetime sexual partners was a risk factor for adult T-cell leukemia/lymphoma. These data support the hypothesis that HAM/TSP is associated with sexually acquired HTLV-I infection, whereas adult T-cell leukemia/lymphoma is not.