The two studies were conducted on volunteer blood donations with a low prevalence of positive antibody. As with other laboratory tests, the lower the prevalance of individuals with antibody present in the sample tested, the higher the number of false positive tests. The combined results demonstrated that positive ELISA tests when repeated are positive only 29 percent of the time, with repeatedly positive tests confirmed by the expensive and unreliable Western Blot test only 23 percent of the time. Two hundred and ninety out of 1,000 would be positive on repeated ELISA testing; of those, only 67 would be confirmed positive by the Western Blot test, and 923 or 92 percent would be false positives. Because AIDS and HTLV-III antibody are low among inmates, false positive results would be common. AIDS and AIDS Related Complex (ARC) diagnoses are made by clinical evaluation and not by the HTLV-III antibody test. Because AIDS can be spread in jails or prisons only through sex or shared needles, this article does not recommend inmate segregation based on test results. Inmate education about needle-sharing and safe sex, as well as providing inmates with condoms, are recommended. 13 references.