Aging is accompanied by a deterioration of physical abilities. For some this limits their mobility and thus their quality of life. Exoskeletons are a class of walking assist device that help reduce the effort required to walk. Currently, powered exoskeletons suffer from short battery life and thus limited usefulness. This thesis presents the design, fabrication, and testing of a novel unpowered ankle exoskeleton to assist normal walking over long distances. The design incorporates a Pneumatic Artificial Muscle (PAM) inflated and used as a passive air spring. To predict the behaviour of the PAM in this distinct application, a distinct dynamic model was developed to include the biaxial stress in the bladder as well as a polytropic gas assumption. Experimental testing was used to validate the model and indicated that the addition of the bladder stress enhanced the performance of the force prediction at low pressure but had negligible impact on the model at higher pressures. The experimental testing also showed that the temperature of the gas inside the PAM varies very slightly during passive elongation cycles, thus, validating an isothermal assumption. Once fabricated, the exoskeleton was tested in human walking trials. I would also like to thank Laurent Levesque as well as my other lab mates for sharing their knowledge and helping to address unexpected problems as they arose.