Although it is natural to regard psychiatric delusions as beliefs, there seem to be significant differences between at least some such delusions and ordinary beliefs. These differences include the comparatively weak influence of psychiatric delusions on the subject’s behavior, emotional life (affect), and “web of beliefs.” They also include the notorious resistance of psychiatric delusions (versus ordinary beliefs) to counterevidence. Such differences have led some psychiatrists and philosophers to speculate that psychiatric delusions may not be genuine beliefs. I take issue with such speculation, noting that the particular features of psychiatric delusions that motivate it, characterize the endorsement of philosophical doctrines whose epistemic status as beliefs is rarely questioned. I then draw attention to a fallacy that has led some theorists to conclude, from the fact that psychiatric delusions are not paradigmatic beliefs, that they are (probably) not genuine beliefs. I call this fallacy the “fallacy of ignoring anomalies.” It occurs whenever one concludes that x is (probably) not a case of y because x is unlike paradigmatic cases of y.