Dapsone, administered at various doses and schedules, has been proven to be a safe and effective alternative to trimethoprim-sulfamethoxazole for prevention of Pneumocystis carinii pneumonia (PCP) in adults with human immunodeficiency virus (HIV) infection. Dapsone is also recommended by the Centers for Disease Control for PCP prophylaxis in HIV-infected children. However, the suggested dosage regimen is based upon clinical experience with children with leprosy and dermatitis herpetiformis rather than pharmacokinetic and pharmacodynamic data obtained from the target patient population. In order to determine a rational dosage regimen that could be tested in clinical studies aimed at the evaluation of dapsone for the prevention of PCP in HIV-infected children, we studied the pharmacokinetics of dapsone following a 2-mg/kg of body weight oral dose in twelve HIV-positive children aged 9 months to 9 years. Plasma was collected at the following times after dapsone administration: 0, 2, 4, 6, 12, 24, 48, 72, and 96 h. The levels of dapsone in plasma were determined by high-performance liquid chromatography. Data were analyzed by noncompartmental methods. Expressed as means +/- standard deviations (ranges), the pharmacokinetic parameters were as follows: peak concentration in plasma, 1.12 +/- 0.48 (0.44 to 1.81) mg/liter; time to peak concentration in plasma, 3.8 +/- 1.3 (2 to 6) h; half-life at elimination phase, 24.2 +/- 7.1 (14.4 to 35.0) h; clearance from plasma divided by bioavailability (CL/F), 1.15 +/- 0.67 (0.37 to 2.63) ml/min/kg; and volume of distribution divided by bioavailability (V/F), 2.25 +/- 1.20 (1.00 to 4.57) liters/kg. Oral CL correlated negatively with age (r = 0.614 and P = 0.034), as did V (r = 0.631 and P = 0.028). As a consequence of the high interindividual variability in growth retardation, pharmacokinetic parameters correlated with measures of body development better than they did with age (e.g., for CL/F to height, r = 0.765 and P = 0.004, and for V/F to height, r = 0.748 and P = 0.005). Since oral CL from plasma and V were positively and highly correlated (r = 0.898 and P = 0.0001), a lower absolute F may be the cause, in part, of higher values for CL/F and V/F in smaller children. The results of this study warrant the testing of a 2-mg/kg dose of dapsone administered twice or thrice weekly to HIV-infected children. The monitoring of drug levels in plasma and dosage adjustment may be necessary for smaller children.