Twenty-six subtotally nephrectomized dogs were used as a model for subclinical renal dysfunction to evaluate the nephrotoxic potential of gentamicin administered according to four different dosage regimens. Dosages were individualized based on the pharmacokinetic disposition of drug in each dog. Gentamicin at 3.75 + 0.15 mg/kg (mean ± standard error of the mean, total daily dose) was given for 12 days in two or three divided daily doses (BID and TID, respectively) or in a 2-h or 4-h once-daily variable-rate infusion (2HI and 4HI, respectively) with loading dose. Analyses of serum chemistries and pharmacokinetic data were performed on the ratio of pretreatment versus posttreatment parameters in individual animals. While serum chemistries and histopathology revealed no significant differences in toxicity between treatment groups, pharmacokinetic analysis revealed a significant difference in the ratio of pretreatment versus posttreatment gentamicin clearance (1.35 ± 0.22, BID; 2.44 ± 0.52, TID; 0.91 ± 0.08, 2HI; 0.91 ± 0.07, 4HI). By using mean population pharmacokinetic parameters (all dogs), predicted times for each treatment group administered 3.75 mg/kg per day to achieve concentrations in serum above the MICs of 2, 4, 6, and 8 ,ug/ml, respectively, were 7.8, 4.2, 2.0, and 0.6 (BID); 6.1, 3.0, 0.5, and 0.2 (TID); 7.1, 5.3, 4.2, and 3.5 (2HI); and 7.4, 5.8, 4.8, and 4.0 (4HI) h daily. This study suggests that decreasing the total daily dosage of drug may decrease the incidence of gentamicininduced nephrotoxicity. Regardless of the dosage regimen, however, regimens may differ significantly in predicted therapeutic efficacy. Predicted 30-min postdosing concentrations in serum were lowest in dogs administered drug TID, and gentamicin clearance decreased in this group with treatment, suggesting that this regimen may be the least efficacious as well as the most prone to causing future nephrotoxicity.The incidence of nephrotoxicity secondary to therapy with gentamicin and other aminoglycosides has been well documented. However, this class of drugs remains an essential therapeutic agent for the treatment of severe gram-negative infections. Traditionally, clinicians have altered doses based on serum creatinine or creatinine clearance in an effort to minimize toxicity. Individuals with compensated subclinical renal dysfunction marked by a decreased number of functional nephrons and normal serum creatinine levels present a diagnostic dilemma, requiring that dosage alteration be based on individual clearance of drug. Decreasing the dosage based on clearance does not preclude toxicity, however, since exposure to drug on a per-nephron basis is increased in patients with reduced functional renal mass (4,8).Previous studies have suggested that some dosage regimens may be less nephrotoxic than others. High peak and high trough concentrations in serum have been implicated as risk factors for development of nephrotoxicity in humans (27). Studies in experimental animals have shown that increasing the dosage interval may be less ne...