The pharmacokinetics of amikacin administered intravenously at currently recommended doses (7.5 mg/kg every 12 h for infants with less than 7 days of life; 7.5 mg/kg every 8 h for infants with greater than 7 days of life) were studied in 28 preterm infants weighing less than 2,500 g (mean + standard deviation, 1.38 0.47 kg; postconceptional age, 30.50 ± 2.86 weeks). The medication was infused over 45 min. Trough and peak serum samples as well as two additional samples were taken at steady state. The results showed a statistically significant inverse relationship between half-life (8.42 ± 2.55 h) and postconceptional age (P = 0.002) and a direct correlation between total body clearance (0.84 ± 0.28 ml/min per kg) and postconceptional age (P = 0.02). These pharmacokinetic data were used to calculate a new dosage schedule for preterm infants. The derived intravenous dosage of amikacin for infants of less than 30 weeks of postconceptional age was 9 mg/kg every 18 h. For infants of greater than 30 weeks of postconceptional age, the dosage was 9 mg/kg every 12 h. Peak and trough levels of amikacin in serum that fell within the therapeutic range were compared by using the currently recommended dosage schedule and the dosage schedule derived from our pharmacokinetic data. There was a reduction in the number of peak and trough levels that fell outside the accepted therapeutic range which was not statistically significant. Extension of the dosing interval and a further increase in the dosage may result in further improvement. Based on these data, the current recommendations are inadequate for the preterm infant. Our derived dosage schedule improved but did not eliminate high trough and low peak levels of amikacin in all infants. The current recommendations should be adjusted for the preterm infant. Ongoing therapeutic drug monitoring is essential to tailor the amikacin dosage to the individual patient.