In inhalation therapy, aerosol deposition studies play a significant role in developing our understanding of the correlation between deposition site in the lungs and clinical effect to the patient. Pulmonary aerosol deposition distribution can be assessed by radionuclide imaging. It can also be predicted using computer-modelling of the particle paths in the airways. Both techniques, however, use anatomical models of the airways that lack three-dimensional (3D) reality, as little information is available on airway topology. Consequently, they cannot relate the aerosol deposition distribution to the subject's own anatomy. The thesis is concerned with the improvement of these methods. An algorithm is developed to obtain the 3D topology and morphology of the airways, from computed tomography (CT) imaging. Realistic 3D airway models are then created by reconstructing the CT image of the airway tree and assigning a different intensity to each airway or to each airway generation. These methods are applied to the CT images of a human tracheobronchial tree cast and of a volunteer. Modelling of realistic 3D pulmonary lobes and segments, based on a region growing process to simulate the "fight for space" that occurs during lung formation, is attempted and applied to the volunteer's magnetic resonance lung images. The aerosol deposition distribution in two radionuclide imaging studies performed on the volunteer is then assessed, using the models representing his own anatomy. The algorithm quickly provides data on the 3D topology and morphology of the airways, from both in vivo and cast analysis, that contribute to the knowledge of the lung anatomy. A non-planar geometry of the airways at bifurcations and clear lobar patterns are shown in the central airways of the datasets studied. The tabulated data and the 3D models of the airways are useful to predict the 3D aerosol deposition and simulate flow patterns in the airways. The study demonstrates that better knowledge of the anatomy allows improved interpretation of the aerosol deposition. A vertical gradient of the concentration of aerosol, increasing from the upper to the lower lobes, is quantified.