Postoperative infection is among the most common complications in patients with cerebrospinal fluid shunt placement. Nafcillin is often used for prophylaxis but not pharmacokinetic data are available perioperatively in pediatric patients. The objectives of this study were to characterize the pharmacokinetics and determine the cerebrospinal concentrations of nafcillin. Ten patients (mean age 8.0 +/- 5.6 years) received three doses of intravenous nafcillin, 50 mg/kg every 6 h; the first dose was administered 1 h prior to surgery. Multiple blood samples were collected during and after surgery and the cerebrospinal fluid sample was obtained at the time of shunt insertion. Urine samples were collected for 24 h after initiation of nafcillin. Nafcillin was analyzed with an HLPC method. The peak serum concentrations ranged from 22 to 107 micrograms/ml; cerebrospinal fluid concentrations ranged from 0.02 to 0.30 (mean 0.16 +/- 0.11) micrograms/ml. The mean total clearance, renal clearance, apparent volume of distribution, and elimination half-life were 0.90 +/- 0.55 l/kg/h, 0.12 +/- 0.04 l/kg/h, 0.70 +/- 0.52 l/kg, and 0.5 +/- 0.1 h, respectively. 16% of total nafcillin dose was excreted in the urine. A 4-fold variability in total clearance and a 10-fold variation in cerebrospinal fluid concentrations of nafcillin was observed in these patients. Further, the concentrations of nafcillin attained in the cerebrospinal do not appear to be adequate, based on its minimum inhibitory concentration of 0.5 micrograms/ml against very susceptible staphylococci. These data, in addition to the fact that an increasing number of staphylococci are becoming resistant to nafcillin, question the usefulness of prophylactic nafcillin in pediatric patients undergoing shunt procedures.