Immunosuppressive therapy is necessary when patients with chronic kidney disease receive a kidney transplant. Certain immunosuppressive agents need therapeutic drug monitoring. The goal of this paper was to identify adaptation rules for Tacrolimus therapy from a clinical data set. For knowledge acquisition, patient data from 1995 to 2008 from the Department of Nephrology and Dialysis of the Vienna General Hospital were used, including patient demographics, laboratory parameters, time since kidney transplantation and other immuno-suppressive drugs administered. Tacrolimus was chosen from the available immunosuppressive drugs. By applying a regression tree, we create homogeneous groups of data. Models were generated for these groups that can predict the level of the drug concentration for the next ward round. A knowledge base was developed on the basis of the determined models, which is used within a clinical decision support system for Tacrolimus therapy planning, which was integrated into clinical routine.