Philosophical theories about the nature of belief can be roughly classified into two groups: those that treat beliefs as occurrent mental states or episodes and those that treat beliefs as dispositions. David Hume's A Treatise of Human Nature seems to contain a classic example of an occurrence theory of belief as he defines 'belief' as 'a lively idea related to or associated with a present impression' (Treatise 1.3.7.5 96). This definition suggests that believing is an occurrent mental state, such as (...) judging, or thinking about something in a particular manner. However, a number of Hume's readers claim to find elements in his writings that are suggestive of a dispositional account of belief. Moreover, these elements are sometimes taken as signs of the inadequacy of Hume's account of belief and his dissatisfaction with it. This paper argues that Hume does hold an occurrence theory of belief, that he has deep-seated reasons for doing so, and that this theory of belief has greater explanatory power than its critics generally allow. I contend in particular that beliefs play an explanatory role in Hume's Science of Man that could not be played by dispositions, but requires that beliefs be introspectible occurrent mental states. Attention to the role of belief in explaining our actions, thoughts, emotional responses, and reasoning reveals that Hume is committed to treating beliefs as occurrent states. (shrink)