This chapter presents the history of research into skill acquisition, and reviews the key questions and theories that have framed this research. Issues include the existence of plateaus in learning curves, the effects of part versus whole task training and massed versus distributed practice, knowledge of results, the form of learning curves and their mathematical description, the power law of practice, transfer of training, and phases of skill acquisition. Theories reviewed are divided up into those that propose that skill acquisition proceeds through a process of strategy refinement (e.g., the theories of Crossman, Anderson (ACT-R), Newell et al. (SOAR), and MacKay, as well as some connectionist theories) as opposed to those that propose that skilled performance results from improved memory retrieval (e.g., the theories of Logan (Instance theory) and Palmeri (EBRW)). The theories are evaluated in terms of their ability to provide accounts versus explanations of the power law of practice.