The drug discovery process is hampered by an extremely poor successful rate (only about 10%) and tremendous costs. The reasons of this dramatic failure rate are drug toxicity and inefficacy issues that are not identified early on by standard in-vitro and in-vivo models. The lung-on-a-chip, which is an advanced in-vitro model, is expected to better predict the effects of respiratory drug candidates in humans and thus reduce the costs. The new model mimics in an unprecedented way the biophysical microenvironment of the air-blood barrier and we could show that the epithelial barrier permeability is significantly affected by the mechanical strain induced by the breathing movements. Therefore, the concentration of inhaled particles (from pollution, aerosols or inhalation drugs) that enters in the blood stream via the lung barrier importantly depends on this dynamic environment. Further we cultured primary human pulmonary alveolar epithelial cells (from patients undergoing a lung resection) on the lung-on-a-chip to better reproduce the in-vivo conditions and to investigate the effect of cyclic strain on the behavior of primary cells.