Reports of Lyme disease from areas where the disease is not endemic have increased. Eighty-six human serum samples from Papua New Guinea (nonendemic for Lyme disease) were examined for the presence of IgG antibodies that recognize Borrelia burgdorferi antigens, using the currently recommended two-tiered system of analysis (sensitive ELISA with Western blot). The percentage of positive tests dropped from 50% to 10% when individual negative controls were included in the two-tiered analysis. Positive serum samples failed to inhibit the growth of B. burgdorferi in culture and did not yield positive reactions in the fluorescent treponemal antibody-absorption test. These characteristics, together with atypical Western blot antigen recognition patterns and the absence of known vectors, provide evidence that seropositive results for these persons are not the result of exposure to B. burgdorferi. Individual negative controls may minimize false-positive results for serologic tests for Lyme disease, and these tests must be interpreted in the context of clinical and epidemiologic data.