Screening for HIV infection can use many algorithms. When two different HIV antibody assays are used, discordant results may occur. To discriminate between HIV seroconversion, HIV variant infection and false positive reactivity, 30 consecutive subjects with two discordant HIV antibody-screening assays were extensively investigated for HIV infection. No subject had HIV seroconversion or reached HIV seropositivity criteria after a follow-up of 3 months. By contrast 36% became HIV negative by the use of both HIV screening assays. p24 Antigen, HIV-1 RNA, HIV-1 DNA, HIV-2 DNA assays and HIV isolation by sensitive culture were unable to identify HIV infection in the 30 subjects with discordant HIV screening assays. The data suggest that the use of two HIV screening assays increase false-positive HIV results without increasing clinical sensitivity. To compliment follow-up of HIV screening, early testing for HIV RNA could be useful to identify or eliminate a recent infection.