The main objective of this paper is to provide an overview and acritical analysis of the recent literature on incorporatinginduced technical change in energy systems models. Special emphasisis put on surveying recent studies aiming at integrating learning-by-doing into bottom-up energy systems models through so-calledlearning curves, and on analyzing the relevance of learning curveanalysis for understanding the process of innovation and technologydiffusion in the energy sector. The survey indicates that this modelwork represents a major advance in energy research, and embedsimportant policy implications, not the least concerning the costand the timing of environmental policies (including carbon emissionconstraints). However, bottom-up energy models with endogenouslearning are also limited in their characterization of technologydiffusion and innovation. While they provide a detailed accountof technical options - which is absent in many top-down models -they also lack important aspects of diffusion behavior that arecaptured in top-down representations. For instance, they fall incapturing strategic technology diffusion behavior in the energysector, and they neglect important general equilibrium impacts(such as the opportunity cost of redirecting R&D support to theenergy sector). For these reasons bottom-up and top-down modelswith induced technical change should not be viewed as substitutes but rather as complements.