In this preliminary study, we evaluated the performance of the Cobas AmpliPrep/Cobas Amplicor HIV-1 Monitor Ultrasensitive Test (CAP; Roche Molecular Systems, Branchburg, NJ) for automated specimen preparation and quantitative detection of human immunodeficiency virus type 1 (HIV-1) RNA and compared it to the Cobas Amplicor HIV-1 Monitor Ultrasensitive Test (MCA; Roche), which includes a manual sample preparation protocol. A dilution panel of a patient sample was prepared. Additionally, 584 EDTA plasma samples were collected from HIV-1 infected patients. Reproducibility was estimated with six assay runs using the dilution panel. The inter-assay coefficient of variation ranged from 39.4 to 48.4% (CAP assay) and from 34.3 to 45.6% (MCA assay), whereas the intra-assay coefficient of variation ranged from 6.2 to 58.0% (CAP assay) and from 4.4 to 57.3% (MCA assay). Comparison of CAP assay results with the HIV-1 copy number of the dilution panel determined by the MCA assay resulted in a good agreement, although the CAP results were found to be slightly lower. A significant correlation between both test systems was found when clinical samples were analyzed. The mean viral load of 152 samples, which were within the linear range of both tests, was 3.70 log(10) HIV-1 copies/ml by the CAP assay compared to 3.73 by the MCA assay. In conclusion, we could demonstrate that the new Cobas AmpliPrep/Cobas Amplicor HIV Monitor Ultrasensitive Test is reproducible and sensitive. In comparison to the assay with manual extraction, no significant difference in HIV-1 RNA copy numbers was observed.