There is uncertainty as to how the effects of antiretroviral treatments on human immunodeficiency virus type 1 (HIV-1) RNA levels and CD4 cell counts can predict reductions in clinical progression to AIDS or death. A meta-analysis was conducted for 27 pairwise comparisons of antiretroviral treatments in 15 randomized trials of antiretroviral treatments. For each trial, three measures of treatment effect were used: (i) 16 week change from baseline in HIV-1 RNA; (ii) 16 week change from baseline in CD4 cell count; and (iii) rate of clinical progression. Treatments which caused greater increases in CD4 cell count and greater reductions in HIV-1 RNA were more effective at reducing the rate of clinical progression (P < 0.05 for each comparison). However, there was variability in the consistency of this correlation between different trials and treatments. The results support the use of both CD4 count and HIV-1 RNA levels as the primary markers of the efficacy of antiretroviral treatment.