Previous studies have shown that passive-elastic exoskeletons with springs in parallel with the ankle can reduce the metabolic cost of walking. We developed and tested the use of an unpowered passive-elastic exoskeleton for walking that stores elastic energy in a spring from knee extension at the end of the leg swing phase, and then releases this energy to assist ankle plantarflexion at the end of the stance phase prior to toe-off. The exoskeleton uses a system of ratchets and pawls to store and return elastic energy through compression and release of metal springs that act in parallel with the knee and ankle, respectively. We hypothesized that, due to the assistance provided by the exoskeleton, net metabolic power would be reduced compared to walking without using an exoskeleton. We compared the net metabolic power required to walk when the exoskeleton only acts at the knee to resist extension at the end of the leg swing phase, to that required to walk when the stored elastic energy from knee extension is released to assist ankle plantarflexion at the end of the stance phase prior to toe-off. Eight (4 M, 4F) subjects walked at 1.25 m/s on a force-measuring treadmill with and without using the exoskeleton while we measured their metabolic rates, ground reaction forces, and center of pressure. We found that when subjects used the exoskeleton with energy stored from knee extension and released for ankle plantarflexion, average net metabolic power was 11% lower than when subjects walked while wearing the exoskeleton with the springs disengaged (p = 0.007), but was 23% higher compared to walking without the exoskeleton (p < 0.0001). The use of a novel passive-elastic exoskeleton that stores and returns energy in parallel with the knee and ankle, respectively, has the potential to improve the metabolic cost of walking. Future studies are needed to optimize the design and elucidate the underlying biomechanical and physiological effects of using an exoskeleton that acts in parallel with the knee and ankle. Moreover, addressing and improving the exoskeletal design by reducing and closely aligning the mass of the exoskeleton could further improve the metabolic cost of walking.