The five available p24 Ag/anti-HIV combined tests were compared to the six third-generation anti-HIV assays mainly used in blood transfusion centers. Among 70 selected HIV-1 positive samples (12 samples from early infected blood donors and 58 from ten commercial panels), 59 were positive with at least one assay. False negative results were observed for zero to six samples with p24 Ag/Ab assays versus seven to 19 with antibody (Ab) tests. In five cases, one or more combined assays gave a positive signal later than the most sensitive Ab screening test. One sample with a high p24 Ag titer was missed by one combined test. The mean time delay between the most sensitive test and the second one was 0.3 to 2 days. The p24 Ag limit of detection was investigated with seven dilutions of the HIV Ag reference. The threshold of the p24 Ag detection was found to be between 65 and 250 pg/mL of HIV Ag. For four of the five combined assays, p24 Ag detectability was assessed with dilutions of infected culture cell supernatants from 13 HIV-1 different genotype strains exhibiting HIV Ag titers from 300 to 450 pg/mL. One of the four combined assays gave negative results but close to the cut-off for three supernatant dilutions (1 B, 1 F, 1 HIV-1/O) and one missed the HIV-1/O dilution. The p24 Ag/Ab combined assays permit an earlier diagnosis of HIV infection than third generation assays even if the yield in terms of reduction of the window period is moderate. They are less sensitive than p24 Ag sreening assays for the detection of this marker. Consequently, the p24 Ag/Ab assays have not been used for the diagnosis of a primary infection instead of p24 Ag screening tests. They must be considered only as good tools for the detection of HIV infection.