Evolutionary robotics–the use of evolutionary algorithms to automate the production of autonomous robots–has been an active area of research for two decades. However, previous work in this domain has been limited by the simplicity of the evolved robots and the task environments within which they are able to succeed. This dissertation aims to address these challenges by developing techniques for evolving more complex robots. Particular focus is given to methods which evolve not only the control policies of manually-designed robots, but instead evolve both the control policy and physical form of the robot. These techniques are presented along with their application to investigating previously unexplored relationships between the complexity of evolving robots and the task environments within which they evolve.