The topic of this paper is the ontology of practical reasons. We draw a critical comparison between two views. According to the first, practical reasons are states of affairs; according to the second, they are propositions. We first isolate and spell out in detail certain objections to the second view that can be found only in embryonic form in the literature - in particular, in the work of Jonathan Dancy. Next, we sketch possible ways in which one might respond to each one of these objections. A careful evaluation of these complaints and responses, we argue, shows that the first view is not as obviously compelling as it is thought by Dancy. Indeed, it turns out that the view that practical reasons are propositions is in no obvious way unworkable and in fact, at least under certain assumptions, explicit considerations can be made in favour of a propositional construal of reasons.