The elucidation of the spread of HTLV-I through high-risk groups and the finite but real incidence of HTLV-I seropositivity in normal blood donor populations in the United States indicates that blood and blood products should be screened for this infectious agent. Because of the ability of the provirus to exist in a quiescent state and the long lag time between exposure and seroconversion, it may be necessary to screen potential blood donors for integrated sequences by gene amplification methodologies in addition to standard serologic testing to protect the blood supply. The detection of the HTLV-I virus often requires multiple modes of testing, even in ATL patients. We have characterized by gene amplification several HTLV-I positive lymphoma patients who were seronegative. We have also identified by radioimmunoprecipitation assays intravenous drugs abusers who have antibody solely to the nuclear pX gene product and who do not, therefore, test positive in an ELISA assay prepared from purified virion proteins. All HTLV-I positive patients need to be counseled about the biohazard status of their body fluids. The fact tha only 1 to 2 per cent of HTLV-I infected persons have any diagnosable disease, coupled with the knowledge that the mean time for the onset of clinical manifestations is some 20 to 30 years following conversion to seropositivity, indicates that this is not a virulent pathogen or a highly transforming virus. These epidemiologic data support the notion of HTLV-I's role as a mitogen or first lesion in a multistep pathway to malignancy. These data are also consistent with the idea of rare random cis-activation of one of many cellular oncogenes following a fortuitous integration event.