To compare frequent measurement with infrequent measurement of human immunodeficiency virus (HIV) RNA levels in the management of antiretroviral therapy, we conducted a clinical strategy study of 206 HIV-infected patients who had <500 CD4 cells/mm(3). Patients were randomized (1.5:1) to undergo frequent monitoring (at baseline and every 2 months) or infrequent monitoring (at baseline and twice yearly), with CD4 cell counts determined every 2 months. Patients received unrestricted antiretroviral therapy. In the primary analysis (at month 6), the frequent group had a mean HIV RNA reduction (+/- standard deviation) of 0.93+/-0.79 log(10) copies/mL, versus 0.48+/-0.83 log(10) copies/mL for the infrequent group (P=.0002). A trend (P=.1) toward improved survival was seen in the frequent group. Given this improved virological response, more frequent HIV RNA measurement than is recommended in published guidelines (every 3-4 months) may be appropriate.