The aim of this thesis was to examine the effects of exercise on gas mixing in the lung during exercise. There were four major stages. Firstly, the existing equipment used with resting subjects was applied to the exercising subject and was found to be inappropriate. Secondly, an in-line system of measuring flow and gas concentration was devised. Thirdly this system was validated with the aid of a physical model and resting subjects. Finally, nitrogen wash-out data were collected from 24 subjects at rest and during progressive exercise at three standard exercise intensities. The dynamic response characteristics of the bag-in-box spirometer at high breathing frequencies (50 min-1) were such that tidal volume was underestimated by almost 50%. The box was too small and its response too a linear for adequate correction factors to be applied. The in-line system, based on a linear relationship between flow and several argon, oxygen, carbon dioxide and nitrogen mixtures ( r = + 0.99, p < 0.01 , Y = 0.2687 FAr + 0.995 ), measured tidal volumes reliably ( CV < 1% ) when expired flow was maintained at 35&deg; C. Thirty-six wash-outs of a 2.4 litre bell jar produced a mean value of 2.461 litres ( SD. 0.034, CV. 1.4% ). The capacity of the in-line system to measure gas mixing efficiency reliably was tested on resting subjects ( six trials each on two days ). Mean values were 76.7% ( SD. 7-5% ) and 76.8% ( SD 4.7% ); mean CV for all trials was 8%. Progressive exercise resulted in significant reduction in lung volume as measured by recovered nitrogen; there was evidence that at the greatest exercise intensity all the nitrogen was not recovered. Decreased diffusion time as a result of greater respiratory frequency may have been responsible. The significantly greater tidal volumes and respiratory frequencies observed on exercise resulted in bigger minute volumes. Both series and alveolar deadspaces increased, but the greater minute volume more than compensated for the growing dead spaces, and so the inspired volume available for mixing was increased. Ventilatory and gas mixing efficiency improved significantly as exercise progressed, but the greatest improvement occurred at the first power output of 50W; thereafter, there was very little change in gas mixing efficiency in spite of three-fold increase in ventilation. It is possible that gas mixing efficiency functions optimally at FRC and that, unlike some other physiological measures, there is little reserve capacity. However, the possibility of gas mixing deficiencies at maximal exercise leading to a ventilatory limit to maximal oxygen uptake remains, and this issue still needs to be investigated.