Imatinib treatment for chronic myeloid leukemia (CML) could be optimized by Therapeutic Drug Monitoring (TDM). The objective of this thesis was to enhance the clinical interpretation of imatinib concentrations measured in the scope of TDM, to improve the evidence of concentration-response relationships, and to evaluate the usefulness of systematic TDM-guided dosage individualization. For this purpose, the population pharmacokinetic (PPK) approach was used and evaluated in several studies. A reference range of imatinib concentrations over the dosage interval was established through a simulation-based meta-analysis of several PPK models. From a validated in-house PPK model, a Bayesian approach for the prediction of the individual trough concentration (Cmin) was developed and validated. Correlations between individual Cmin and treatment response were confirmed in a large European patient cohort study. The clinical usefulness of systematic TDM was evaluated through a prospective randomized controlled trial. The Bayesian approach allowed easing the sampling logistics in the trial.