The main aim of the current chapter is to situate automatic imitation in the broader historical context of research on the functional mechanisms underlying imitation. Intuitively, imitation is a very simple act: you do what you see. However, on closer inspection, the question arises as to how a perceptual representation of a movement can be transformed into a corresponding motor program (the so-called correspondence problem of imitation). Research on the correspondence problem has a long history in psychology. Three major theoretical approaches to solve the problem have been proposed: first, imitation has been conceptualized as an innate mechanism; second, from a learning theoretical perspective, imitation has been understood as the result of simple learning mechanism, equating imitation with any other learned behaviour. Finally, ideomotor theory has conceptualized imitation as the result of ideomotor learning distinguishing it from other learned behaviour. While automatic imitation was originally developed in the context of ideomotor theory, it has also been used to support learning theoretical approaches to the correspondence problem. Almost 25 years of research on automatic imitation have not solved the correspondence problem but have provided cognitive psychology with a valuable tool to experimentally investigate imitative behaviour and to bridge the gap between cognitive and social-psychological approaches to imitation.