Cyclosporine is one of the most widely used immunosuppressive agents in organ transplantation. Due to large inter- and intra-individual variations, its behavior in the specific patient is still difficult to predict. Dosage optimization is thus mainly performed on a trial-and-error basis. In this paper, we present a new program based on the population kinetics approach, which was designed to help physicians in the difficult task of adjusting patient specific cyclosporine dosing regimens. Dose optimization is carried out by model simulation, using a two-compartment mathematical model of cyclosporine kinetics to predict the drug behavior in the patient. Two of the model parameters are assumed from the literature, the other two are estimated from the patient data through a Bayesian estimation procedure. Previous information needed by the Bayesian algorithm is derived by a population analysis, performed beforehand and based on a nonlinear mixed effect model. A user-friendly graphical interface written in Delphi under Windows makes the program easily accessible to physicians. A preliminary retrospective validation of the program, performed on data from 18 renal transplanted patients, yielded very satisfactory results.