Out of all parameters used to describe gait, overground speed is one of the most important. The importance of gait speed is highlighted when used as a measure of performance during exercise, or as a measure of function when walking ability is compromised. Because the ability to control gait speed is imperative to reach optimal results in both exercise and gait rehabilitation, a system that helps people to control their overground speed more accurately might be beneficial. Developing an overground speed control system was the main goal of this thesis. To gain insight in the performance enhancing effects that can be expected from such a system, my colleagues and I first determined the ability of recreational runners to accurately control their own speed. We then used a simulation approach to estimate the effect of pacing inaccuracy on optimal running performance. Our simulation results suggested that the existing pacing error (2.3±4.6%) would decrease optimal performance by approximately 5% for an average recreational runner. These results indicate that the performance of recreational runners could be improved by minutes for typical race distances, simply by helping them achieve and maintain their optimal speed. To determine the viability of controlling overground speed by prescribing step frequency, we quantified the dynamic response in walking and running speed following controlled perturbations in prescribed metronome frequency. We found that perturbations in metronome frequency triggered rapid and predictable changes in speed, suggesting that overground speed is indeed controllable by prescribing step frequency. However, due to the variability present in the speed response, both within and between individuals, accurately controlling overground speed using an open-loop speed control system is not possible. To improve speed control performance we developed and built a closed-loop speed control system, which made the metronome frequency directly dependent on the instantaneous speed error. We tested the performance of this system in both walking and running, and found that the speed control accuracy of a closed-loop system was significantly better compared to self-paced running and an open-loop speed control system. Finally, we translated the speed control system into a training tool available to the general public.