The subject of this thesis is the application of the dynamic walking approach to human gait. This work is motivated by the needs of persons with disabilities and by a desire to expand basic understanding of human walking. We address human gait from the perspective of dynamic walking, a theoretical approach to legged locomotion which emphasizes the use of simple dynamical models and focuses on behavior over the course of many steps rather than within a single step. We build on results from prior dynamic walking research and develop new areas of exploration, with energetics and stability providing context. We focus on three areas:​ improvement of prosthetic foot design, the function of arm swinging, and evaluation of balance among the elderly. These issues are addressed by use of dynamic walking models and controlled human subject experiments. We propose a Controlled Energy Storage and Return (CESR) foot prosthesis to increase push-off work and reduce energy expenditure in amputees, and tested a prototype experimentally. To better understand the role of arms swinging in gait, we developed a simple dynamic walking model with free-swinging arms and performed human subject experiments in which subjects swung their arms in various ways. Finally, we studied the effects of aging on balance during walking using a computational model and a human subject experiment in which younger and older adults walked overground for hundreds of consecutive steps. These models and experiments each expand our understanding of the fundamentals of gait and indicate pathways toward assisting individuals with disabilities. Taken as a whole, this work emphasizes the utility of the dynamic walking approach.