Serum samples collected in 1984 during a malariometric survey of two villages in the East Sepik Province of Papua New Guinea were tested for antibodies to HTLV-I. None of the villagers showed any symptoms suggestive of retrovirus infection. Eighteen of the 186 (9.5%) sera tested at that time were found to be positive. Blood samples were subsequently obtained from fifteen of the eighteen positives and subjected to analysis by enzyme-linked immunosorbent assay (ELISA), radioimmuno assay (RIA), radioimmunoprecipitation assay (RIPA), and Western blot (WB). Fourteen of the fifteen gave a positive ELISA response, but none were unequivocally positive by p 24 RIA. All sera tested were reactive to gag antigens by WB, but gave "indeterminate" results currently accepted criteria. Notably absent from the WB profiles of all of the study subjects was an antibody response to HTLV-I envelope protein gp 46. It is possible that these antibody responses are directed against a variant of HTLV-I or to a novel retrovirus which possesses core antigens similar to those of HTLV-I but has different envelope antigens. Until a virus is isolated, or the viral genome is identified in infected lymphocytes, the possibility remains that the response may be due to factors unrelated to retrovirus infection.