Plasma HIV-1 RNA was measured in 306 samples, collected from 273 highly active antiretroviral therapy-experienced men, using both the Roche COBAS TaqMan (limit of detection = 20 copies per milliliter) and Roche Amplicor (limit of detection = 50 copies per milliliter) assays. Mixtures of Gaussian distributions incorporating left-censored data were used in analyses. The more sensitive TaqMan assay estimated that 23% and 0.0003% of HIV-1 RNA values would be below 1 copy per milliliter and 1 copy per 3 L, respectively. This is in sharp contrast to the overestimation provided by the less sensitive Amplicor assay, whereby the corresponding predicted percentages were 51% and 1%. Both assays appropriately characterized suboptimal virologic response as the rightmost peaks of both distributions provided an excellent fit to the observed data. Our results based on a widely available 20 copies per milliliter sensitive assay reproduce those obtained using customized assays that quantified HIV-1 RNA values as low as 1 copy per milliliter.