Newborn infants cared for in neonatal intensive care units may develop nosocomial infections. Cefepime, a "fourth-generation" cephalosporin (i.e., with activity against virtually all of the chromosomal-beta-lactamase-producing and many extended-spectrum-beta-lactamase-producing organisms), provides excellent activity against many gram-negative pathogens resistant to expanded-spectrum cephalosporins currently used to treat neonatal infections. The purpose of this study was to determine the pharmacokinetics of cefepime in this population to optimize dosing and minimize potential adverse events. Premature and term infants <4 months of age hospitalized in two neonatal intensive care units were studied. Limited pharmacokinetic (PK) sampling occurred following a dose of cefepime at 50 mg/kg of body weight infused over 30 min. Population pharmacokinetic parameters were determined using the program NONMEM. Fifty-five infants were enrolled. Their average (+/- standard deviation) gestational age at birth was 30.5 +/- 5.3 weeks, and their average postnatal age at PK evaluation was 14.5 +/- 14.7 days. In the final PK model, cefepime clearance (CL) was strongly associated with serum creatinine (SCr) (CL [ml/min/kg] = 0.26 + 0.59/SCr). The volume of distribution for infants with a postconceptional age of <30 weeks was larger than that for infants with a postconceptional age of >30 weeks (0.51 versus 0.39 liter/kg, respectively). The Bayesian analysis-predicted cefepime trough concentration at a dose of 50 mg/kg every 12 h for infants < or = 14 days of age was 29.9 +/- 16.6 microg/ml. Cefepime, dosed at 30 mg/kg/dose every 12 h for infants less than 14 days of age, regardless of gestational age, should provide antibiotic exposure equivalent to or greater than 50 mg/kg every 8 h in older infants and children.