Six procedures for quantifying plasma human immunodeficiency virus type 1 (HIV-1) RNA were evaluated by nine laboratories. The procedures differed in their sample volume and preparation of samples and methods of amplification and detection. Coded samples in a IO-folddilution series of HIV-L-spiked plasma were correctly ranked by all six procedures. Subsequently, coded duplicate plasma samples from 16 HIV-I-infected patients were tested using a common set of standards. Several HIV-1 RNA procedures were sufficiently reproducible so that an empiric 4-fold change could be viewed as significant. HIV-I RNA levels in the patients (up to 370, 000 RNA copiesjmL) correlated with proviral HIV-I DNA and were inversely correlated with CD4 cell counts; HIV-1 RNA assays were more sensitive than plasma viremia, standard p24 antigen, or immune complex-dissociated p24 antigen assays. This study demonstrated that several HIV-1 RNA quantitative assays are ready for use in clinical trials.