Past research suggests that a categorical event is perceived to be more likely if its subcases are explicitly delineated or "unpacked." In 6 studies, we find that unpacking can often make an event seem less likely, especially when the details being unpacked are already highly accessible. Process evidence shows that the provision of greater detail accompanying unpacking reduces the simplicity of an event and that this dysfluency is used as a negative cue for likelihood. This work establishes processing fluency as a mechanism that opposes the other effects of unpacking, such as enhanced accessibility. (Contains 1 table, 2 figures and 3 footnotes.)