The indinavir dosage regimen currently used for human immunodeficiency virus (HIV)-infected children is not based on pharmacokinetic data obtained in the target patient population. The purpose of our study was to characterize indinavir pharmacokinetics and pharmacodynamics in HIV-infected children. Eleven children (age range, 9.0 to 13.6 years; weight range, 21.7 to 56.0 kg) receiving indinavir (500 mg/m(2) every 8 h) in combination with lamivudine and stavudine were studied. The correlation of indinavir pharmacokinetic parameters and demographic parameters was evaluated. Also, the pharmacodynamic relationship between parameters of indinavir exposure and parameters of renal toxicity and immunologic recovery was studied. The area under the indinavir concentration-time curve (AUC) and patient body surface area (BSA) showed a significant negative correlation (r = 0.73; P = 0.012). Patients with smaller BSA had excessive indinavir AUC compared to adults. On the other hand, the median minimum drug concentration in plasma (C(min)) was lower than that reported for adults. The maximum indinavir concentration in serum was higher in patients with renal toxicity (5 out of 11 children), but the difference was not statistically significant (15.3 +/- 8.2 versus 9.8 +/- 4.4 mg/liter; P = 0.19). There was a trend toward higher immunologic efficacy in patients with greater indinavir exposure: the time-averaged AUC of the percentage of CD4(+) lymphocytes over the baseline value for patients with indinavir C(min) > 95% inhibitory concentration (IC(95)) was higher than in patients with C(min) < IC(95) (P = 0. 068). Our study suggests that a dose reduction may be appropriate for children with small BSA and that a 6-h dosage regimen may be indicated for a substantial percentage of patients. Due to the low number of patients enrolled in this study, our results should be confirmed by a larger study.