OBJECTIVE: To investigate reversible metabolic pharmacokinetics of prednisone and prednisolone in plasma and blood nucleated cells in 10 healthy male volunteers. METHODS: Each subject participated in a three-phase randomed crossover study. The three phases were (1) no drug administration, (2) oral administration of a single 50-mg prednisone, and (3) oral administration of a single 50-mg prednisolone. The plasma and nucleated cells were separated from blood for HPLC analysis of prednisone, prednisolone and cortisol. RESULTS: After administration of prednisone, The maximum plasma concentrations (Cmax) of prednisone and prednisolone were 43.0±11.8 and 443.8±198.3 ?g.L-1, respectively. The areas under concentration-time curve (AUC) of prednisone and prednisolone were 382±127 and 2823±943?g.h.L-1, respectively. After administration of prednisolone, The Cmax of prednisone and prednisolone were 47.2±14.0 and 538.9 ±92.4?g.L-1, respectively. The AUC of prednisone and prednisolone were 335±120 and 3664±1170?g.h.L-1, respectively. There were statistically significant differences in AUC ratio of prednisolone to prednisone for both doses (7.44±0.64 vs. 11.01±0.57, P 0.01). Other pharmacokinetic parameters, such as volume of distribution (Vd), clearance (CL), elimination half-life (T1/2), were calculated according to a reversible metabolic non-compartment model. It appears from these parameter values that the reversible metabolism and other eliminations of prednisone are faster than those of prednisolone. The fraction recycled through interconversion (RF) was 0.68±0.08, After administration of prednisone, the Cmax of prednisone and prednisolone in blood nucleated cells were 0.90±0.27 and 8.93±1.78 ?g.10-9 cells, respectively. After administration of prednisolone, the Cmax of prednisone and prednisolone in cells were 0.83±0.32 and 9.26±3.71?g.10-9 cells, respectively. There was a significant correlation between plasma concentrations and blood nucleated cells concentrations, which suggested prednisone and prednisolone transfer into cells by passive diffusion. The of prednisolone in cells (2.11±0.67 and 2.57±0.35 h, after administration of prednisone and prednisolone respectively) were significantly (P 0.01) shorter than either of prednisone in cells (4.28±1.03 and 4.20±0.95 h, after administration of prednisone and prednisolone respectively) or of prednisolone in plasma (5.31±2.26 h). CONCLUSION: The interconversion between prednisone and prednisolone might be a nonlinear process and it is the most primary elimination process in human as well. The pharmacokinetics of prednisolone in cells is not fully consistent with that in plasma.