The effect of variations in plasma protein binding on the renal excretion of cefonicid was assessed by using isolated perfused rat kidneys. Cefonicid exhibits preferential binding ex vivo to human serum albumin (HSA), as opposed to bovine serum albumin (BSA), and is eliminated mainly by tubular secretion, a process that was reported to be dependent on the total drug concentration. This contradicts previous studies with antimicrobial compounds and other drugs of low renal extraction in which the unbound drug concentration was shown to be the driving force for carrier-mediated tubular transport. To clarify this discrepancy, we performed perfusion studies by using 6% BSA at initial concentrations of 200 micrograms/ml (n = 6) and 20 micrograms/ml (n = 9) and in a combination of 4% BSA plus 2% HSA at initial concentrations of 200 micrograms/ml (n = 4). The excretion ratio [ER = CLR/(fu x GFR)] of cefonicid decreased with increasing unbound concentrations, whereas no apparent relationship with the total concentration was evident. At similar total concentrations of cefonicid, the renal clearance remained unchanged; the secretion clearance increased significantly in the 4% BSA-2% HSA experiments, reflecting the reduced unbound fraction and unbound drug concentration of cefonicid. The excretion ratio data were compatible with a model in which Michaelis-Menten kinetics were required to describe active transport and secretion was dependent on the unbound cefonicid concentration. As a result, changes in plasma protein binding as a result of drug interactions or disease states could significantly influence the tubular transport capability of compounds with low renal extraction.