To evaluate the risk of transfusion-related transmission of HTLV-I in Jamaica, a prospective study was initiated, prior to availability of a licensed HTLV-I serological screening assay. This information would prove useful in formulating strategies for blood-donor screening. We followed 118 pre-transfusion HTLV-l-negative transfusion recipients at monthly intervals posttransfusion for I year. Laboratory and questionnaire data were obtained at each visit to evaluate the clinical and immunological status of recipients. Cumulative incidence of HTLV-I seroconversion was estimated and risk-factor data associated with seroconversion among 66 HTLV-l-exposed transfusion recipients were analyzed. Seroconversion occurred in 24/54 (44%) of recipients of HTLV-l-positive cellular blood components, 0/12 recipients of positive non-cellular donor units and 0/52 recipients of HTLV-l-negative donor units. Significant risk factors associated with recipient seroconversion were receipt of a seropositive cellular blood component stored for less than one week [odds ratio (OR) = 6.34, 95% confidence interval (CI) = 1.83 to 21.92], male sex (OR = 4.79, 95% CI = 1.15 to 20.0) or use of immunosuppressive therapy at time of transfusion (OR = 12.20, 95% CI = 0.95 to 156). Risk of blood-borne infection per person per year in Jamaica was estimated to be 0.009%. Our results confirm that blood transfusion carries a significant risk of HTLV-I transmission and that screening of donor blood effectively prevents HTLV-I seroconversion. Recipients at greatest risk for seroconversion were those who required multiple transfusions or who were receiving immunosuppressive therapy at the time of transfusion. These patients should be given priority in receiving selectively screened blood components, if universal blood-donor screening for HTLV-I is not possible.