Two doses of gentamicin (2 and 7 mg/kg of body weight) were administered to 11 healthy volunteers in a randomized, crossover single-dose study to compare their pharmacokinetics. Doses were infused over 1 h with a syringe infusion pump, and 14 concentrations in sera were obtained over an 8-h period. Concentration in serum versus time data were fitted to a two-compartment pharmacokinetic model. In addition, to mimic the clinical setting, subjects' data were fitted by the Sawchuk-Zaske method. Distributional and postdistributional peak concentrations, along with the last obtained concentration in serum, were utilized to compare the following pharmacokinetic variables: volume of distribution at steady state (Vss), half-life, clearance (CL), and maximum concentration in serum (Cmax). With two-compartment pharmacokinetic fitting, significant differences in distribution half-life (average, 21.8 and 41.6 min [P < or = 0.05]) and gentamicin CL (76.6 +/- 6.6 and 67.2 +/- 4.2 ml/min/1.73 m2 [P < or = 0.001]) were found between traditional-dose and high-dose groups, respectively. When the data for concentrations in sera were fitted to a one-compartment pharmacokinetic model by using either the distributional or the postdistributional Cmax, statistically significant differences (P < or = 0.001) were found between Vss, half-life, CL, and Cmax values for both dosage groups. The results show that the pharmacokinetics of gentamicin at a large dose differ significantly from those at the traditional dose. This information has direct implications for once-daily aminoglycoside (ODA) literature when the Cmax values reported are distributional and therefore show falsely high Cmax/MIC ratio estimates. In addition, ODA nomogram dosing tools developed with distributional Cmax values are probably inaccurate.