It was shown previously that about 97% of the anti-HIV-1 group O strain-positive samples were detected by crossreaction with native HIV-1 gp160 (Van Binsbergen et al., Evaluation of a new third generation anti-HIV-1/anti-HIV-2 assay with increased sensitivity for HIV-1 group O, J. Virol. Methods 60 (1996) 131-137). Fourteen out of 17 new anti-HIV-1 group O positive samples, selected with the Enzygnost HIV-1/2 plus assay, were already reactive when tested with HIV-1 gp160. When tested by the Vironostika HIV Uni-Form II plus O microELISA all 17 samples were reactive, demonstrating the necessity to implement an HIV-1 group O-specific antigen in the assay. On the other hand, it was surprisingly found that 40 out of 43 (93%) of anti-HIV-1 group M-positive samples, belonging to strain A, B, C, D, E or F, were detected by crossreaction with the HIV-1 group O (strain ANT70) synthetic peptide incorporated in the Vironostika HIV Uni-Form II plus O. Only HIV-1 subtype D-positive samples did not react with this peptide, presumably because of the presence of a histidine residue in the immunodominant region of HIV-1 subtype D gp41. Both crossreactions make the Vironostika HIV Uni-Form II plus O microELISA also sensitive for anti-HIV-1-positive samples originating from different geographical regions and resulting from different HIV-1 subtype infections. With an unusual seroconversion panel in which p24 Ag was present persistently, many anti-HIV-1/-2 assays produce alternating positive/negative results in anti-HIV antibody-positive bleeds. It was shown that the use of viral p24 and gp160 in a direct sandwich, allowing detection of anti-HIV IgG and IgM, explains the identification of all anti-HIV-positive bleeds by the Vironostika HIV Uni-Form II plus O. The high sensitivity of the plus O assay was confirmed with clinical samples of a so-called anti-HIV-1 low titer panel. The specificity of the Vironostika HIV Uni-Form II plus O determined in five blood transfusion centers, based on 135070 tests, was 99.97%.