This paper presents an engineering approach for optimal drug delivery to the human brain. The hierarchical design procedure addresses three major challenges: (i) physiologically consistent geometric models of the brain anatomy, (ii) discovery of unknown transport and metabolic reaction rates of therapeutic drugs by problem inversion, and (iii) a rigorous method for determining optimal parameters for delivering therapeutic agents to desired target anatomy in the brain. The proposed interdisciplinary approach integrates medical imaging and diagnosis with systems biology and engineering optimization in order to better quantify transport and reaction phenomena in the brain in vivo. It will enhance the knowledge gained from clinical data by combining advanced imaging techniques with large scale optimization of distributed systems. The new procedure will allow physicians and scientists to design and optimize invasive drug delivery techniques systematically based on in vivo drug distribution data and rigorous first principles models. (C) 2007 Elsevier Ltd. All rights reserved.