This paper asks what it means to say that a disorder is a “real” disorder and then considers whether culture-bound syndromes are real disorders. Following J.L. Austin I note that when we ask whether some supposed culture-bound syndrome is a real disorder we should start by specifying what possible alternatives we have in mind. We might be asking whether the reported behaviours genuinely occur, that is, whether the culture-bound syndrome is a genuine phenomenon as opposed to a myth. We might be wondering whether the condition should rightly be considered a disorder, as opposed to some sort of non-disorder condition (for example, a non-disorder form of deviance, or a potentially valuable condition). We might want to know whether the culture-bound syndrome is really a distinct disorder, in the sense that scientific classification systems should include it as a separate category, or whether it is just a variant of a universally occurring disorder. I argue that some specific difficulties can arise with determining whether a culture-bound syndrome is a real disorder in each of these three senses. However, the frequent assumption that real disorders will necessarily occur universally, and that those that occur only in certain environments are suspicious is not generally justified.