A quantitative cell microculture assay (QMC) was used to measure the human immunodeficiency virus type 1 (HIV-1) peripheral blood mononuclear cell (PBMC)-associated titer in 109 subjects rolled in an open-label phase I/II study of didanosine monotherapy or combination therapy with zidovudine. The titer was inversely correlated with CD4+ cell count at baseline (r = .37, P = .001). After 12 weeks of therapy, subjects showed a significant decreases in virus titer and those with the highest baseline virus titers had the greatest increase in CD4+ cell number (r = .430, P = .002). The QMC assay was more sensitive (98%) for assessing the antiretroviral effect of therapy than was immune complex-dissociated HIV p24 antigen (32%) or plasma culture (3.4%). Estimated sample sizes for phase I/II clinical trials were derived using the within-subject QMC SD of .72 log10 infectious units per 10(6) PMBC.