Indinavir plasma levels are associated with antiretroviral efficacy; however, little data are available regarding toxicity. We assessed the relationship between indinavir pharmacokinetic (PK) characteristics and severe nephrolithiasis as well as other severe or serious adverse reactions. Patients included in the ANRS CO8 APROCO-COPILOTE cohort and receiving 800 mg indinavir three times daily as a first-line protease inhibitor were eligible for this study. To be included in the analysis, their plasma sample at month 1 (M1) had to be available (n = 282) to estimate using population PK modeling, indinavir PK characteristics, ie, maximum (Cmax) and trough plasma (Cres) concentrations, area under the curve (AUC), and observed/predicted concentration ratio (CR). A Cox model was used to estimate the independent effect of Cmax, Cres, AUC, and CR on the hazard of severe nephrolithiasis and serious adverse reactions. At M1, median Cmax was 6205 ng/mL, Cres 631 ng/mL, AUC 24,242 ng . h/mL, and CR 0.6. After a median follow up of 12 months, 11% of patients (30 of 282) had experienced at least one serious adverse reaction among which 12 were nephrolithiasis. In the multivariate analyses, early high indinavir Cres (ie, >/=1000 ng/mL at M1) was associated with a higher rate of severe nephrolithiasis (hazard ratio = 6.7; 95% confidence interval = 1.8-25.2; P < 0.01) and was also associated with a higher rate of all serious adverse reactions but only when nephrolithiasis were included among those cases. Prospective and early indinavir Cres determination should be recommended in the patient's care management and dosage adjustments.