Are the eating disorders discrete diagnostic entities or do they fall along one or more continua ranging from normal body weight, eating behavior, and weight concerns to severely disturbed patterns? Researchers have debated this question for at least 30 years and have used numerous creative strategies to examine this and related questions. This body of research is reviewed with particular attention to the more recent use of taxometric methods. Although somewhat mixed, much of the earlier research has been interpreted as supporting the continuity model. However, more recent taxometric research suggests the presence of one or more latent discontinuities, particularly with disorders associated with binge eating. These findings have implications for assessment, treatment, and possible prevention of eating disorders, and may ultimately allow us to better predict who will or will not develop an eating disorder in response to dieting, as well as who will or will not respond to particular treatments for an existing eating disorder.