By mirroring the success of biological systems, legged robots have the potential to be successful in almost every terrestrial environment. While legged machines have made significant advancements in the past 20 years, there still exists a considerable gap between what they can achieve and the abilities of animals. In this talk I’ll discuss some of my recent work on exploring how internal degrees of freedom – spines and tails – can help to create more agile robots. First I’ll discuss my recent results showing how an internal degree of freedom can enable gravity to energize a simple hopping robot using a novel stepping strategy. I’ll present my formal analysis of the system which explains how the strategy controls the distribution of energy in an intuitive manner and present simulation and hardware results showing our robot hopping at speeds of up to 8.85 leg lengths per second. Next I’ll present some new work where I show how a coupled oscillator model, traditionally used for studying vibrations, can explain the role of a spine in trotting and bounding.