Publisher Summary This chapter reviews the results of bidirectional control experiments, and then considers in detail whether they provide evidence of imitation. However, the purpose of the discussion is not to insist that the bidirectional control evidence of imitation is completely secure. In the course of the last century, most, if not all, putative demonstrations of imitation in animals are challenged usually with good reason. Therefore, no one with an eye to history could be confident that recent studies are decisive. Rather, the present discussion uses data from the bidirectional control experiments as a test case to examine how and why imitation is distinguished from other kinds of social learning. However, the results of two subsequent experiments suggested that, in the bidirectional control procedure, it is the demonstrators' action that influences the observers' behavior. Thus, the bidirectional control effect does not seem to belong to any established category of non-imitative social learning. Although, in practice, imitation is defined largely by exclusion, as a default category, there have been some influential positive definitions, attempts to say what imitation is, rather than what it is not. To summarize, the foregoing discussion suggests that the bidirectional control effect does not belong to any established category of non-imitative social learning.