Prosthetic feet are limited in their ability to mimic the energy-recycling behavior of an intact ankle, negatively affecting lower-limb amputees’ gait in terms of metabolic cost and walking speed. To overcome these weaknesses, a novel prosthetic ankle based on hydraulics is described here. The ankle joint drives two cams, which in turn drive two hydraulic rams. One cam-ram system captures the negative work done from foot-flat until maximum dorsiflexion, by pumping oil into an accumulator, while the other returns positive work during push-off providing forward propulsion through fluid flowing from the accumulator to the ram. Simulation results are promising: of the total negative work done by the prosthetic ankle over the gait cycle (i.e., the maximum amount of energy available to be stored), 78% is returned, mainly during push-off; 14% is carried forward for future gait cycles; and 8% is lost. The estimated prosthesis height and mass are approximately <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${26}.{5} {cm}$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${2}.{3} {kg}$ </tex-math></inline-formula> . Nonetheless, further work is necessary to realise a prototype for bench and in-vivo testing. By mimicking intact ankle torque and efficiently storing and returning energy at the ankle joint, this new design may contribute to reducing amputees’ metabolic cost of walking.