When gas is inspired into the lungs it mixes with the gas left in the lungs at the end of the previous expiration. In health, this mixing is relatively efficient but in disease this gas mixing is impaired. Impairment of mixing is brought about by two processes, a maldistribution of inspired gas into a parallel system, or an inhibition of diffusion. The aim of this thesis was to illuminate the contributions made by these two processes to this impairment. This was done by investigating single and multi-breath alveolar mixing efficiency (AME) in never smokers and patients with chronic airway disease and by the use of mathematical models. Both indices of alveolar mixing efficiencies were found to be reproducible in never smokers and patients, AME(mb) and AME(sb) were both significantly lower in patients than in never smokers. A two-compartment parallel model was devised to evaluate the contribution of regional inhomogeneity. This showed that the only way both values of AME(sb) and AME(mb) could be reproduced was to include a component of diffusion limitation. Another model was devised to evaluate the contribution of diffusion limitation to mixing impairment. This model proved to be inadequate to permit any conclusions. The work presented in this thesis suggests that gas mixing impairment results from a maldistribution of inspired gas and gaseous diffusion limitation with the latter playing the most significant role in the generation of the alveolar slope, more especially in diseased lungs.