Study Objectives. To determine the variability of indinavir pharmacokinetics in patients attending an outpatient clinic, and to explore relationships between indinavir exposure and antiviral effect. Design. Open, formal pharmacokinetic evaluation. Setting. University-affiliated clinical research center. Patients. Forty-three adults infected with the human immunodeficiency virus (HIV) receiving therapy with indinavir and concomitant nucleoside reverse transcriptase inhibitors. Intervention. Indinavir concentrations were measured after patients were observed taking an 800-mg oral dose, and pharmacokinetic parameters were determined using a one- compartment oral absorption model. Virologic and pharmacologic characteristics were compared in a subset of 23 patients who were protease inhibitor naive before receiving indinavir. Measurements and Main Results. Mean indinavir pharmacokinetics were similar to those reported previously significant intersubject variability in systemic exposure was observed in patients receiving the same dosage; the 8-hour area under the curve (AUC8) ranged from 5.4-68.0 μM · hour. In protease inhibitor-naive subjects, the indinavir AUC8 was statistically higher in those with undetectable plasma HIV RNA (30.7 μM · hr) versus detectable plasma HIV RNA (22.4 μM · hr, p=0.035). Measured concentrations 5 hours after the dose and extrapolated 8- hour concentrations were also significantly higher in patients with undetectable plasma HIV RNA (both p=0.007). Conclusions. Indinavir plasma concentrations were highly variable among patients receiving the same dosage. Patients with an undetectable plasma HIV RNA level who were protease inhibitor naive had statistically higher indinavir concentrations and slower oral clearance than the group with detectable HIV RNA. Relationships between indinavir concentrations and anti-HIV effect provide a basis for quantifying the pharmacologic contribution to the heterogeneity in therapeutic response.