<p> The motor driven treadmill is often used in research as a convenient tool for simulating overground running. There has been varied opinion in the literature regarding the accuracy of this assumption. The major difference that has been quantified is the variation in treadmill belt speed as a result of the forces applied by a runner. In comparison, the earth does not vary its speed during overground running. The aim of the present study was to more clearly define the causes of treadmill belt-speed variation and to elucidate its effects on running mechanics. </p> <p> An in-lab fabricated tachometer was used to determine accurate treadmill belt speed while the treadmill was challenged by five subjects weighing 55.2 to 99.6 kg running at four speeds of 2.6, 3.1, 3.5 and 4.0 m/s. The actual running velocity was found on average to be 0.62% higher than the treadmill display setting. The intra-step belt-speed variation ranged from 4.2 to 8.6 % of average belt velocity. Linear regression analysis showed that 86 % of the variance in intra-step belt-speed variation was attributed to total body mass and a further 10 % attributed to running speed. </p> <p> The effect that this variation had on running mechanics was determined from the power transfer between the foot and belt, as calculated from the product of the change in belt speed and the horizontal ground reaction force. The horizontal force, as calculated using a segmental acceleration approach, did not show complete agreement with simultaneously recorded forceplate data. It was found that an average of 4.49 J flowed to the treadmill during the eccentric phase of running and 3.37 J of energy flowed to the runner during the concentric phase of running. Despite inaccuracies in the calculation, the mathematical approach used in this study permitted insight into the theoretical benefit of belt-speed variation in treadmill running. </p> / Thesis / Candidate in Philosophy