Six healthy male volunteers were given 5, 20, and 50 mg of oral prednisone and 5, 20, and 400 mg doses of intravenous prednisolone. Plasma and urine concentrations of prednisone and prednisolone were determined by HPLC, and the binding of prednisolone to plasma proteins was measured by radioisotopic and equilibrium dialysis techniques. The pharmacokinetics of both oral prednisone and intravenous prednisolone were dose-dependent. The mean oral dose plasma clearances of prednisone ranged from 572 ml/min/1.73 m 2 for the 5 mg dose to 2271 ml/min/1.73 m 2 for the 50 mg dose. Changes in prednisone half-life were insignificant, but increases in the half-life of its metabolite were dose-dependent. The systemic plasma clearance of i.v. prednisolone was dose-dependent and increased from 111 to 194 ml/min/1.73 m 2 over the 5 to 40 mg i.v. dosage range. The steady-state volume of distribution also increased, but little change in mean transit time and half-life was found. The binding of prednisolone to plasma proteins was markedly concentration-dependent, and a two compartment, nonlinear equation was used to characterize the effective binding of prednisolone to transcortin and albumin. The apparent pharmacokinetic parameters of protein-free and transcortin-free prednisolone were relatively constant with dose. The interconversion of prednisone and prednisolone varied with time and dose, although prednisolone concentrations dominated by 4- to 10-fold over prednisone. In urine, 2-5% of either administered drug was excreted as prednisone and 11-24% as prednisolone. The apparent renal clearances of both steroids were also nonlinear and unrelated to protein binding. These studies indicate that the pharmacokinetics of prednisone and prednisolone are dose-dependent and that protein binding does not fully explain their apparent nonlinear distribution and disposition.