Bayle's conception of reason is notoriously difficult to unravel, as are its consequences for the rationality of religious belief. The secondary literature has generally coalesced around two interpretations of Bayle's conception of reason. The “superskeptical” interpretation holds that reason is the source of its own undoing, not to be trusted; religious belief turns out to be irrational on this conception of reason, but this is hardly cause for alarm. The jusqu'au bout (to the very end) interpretation holds that reason is (...) to be followed wherever it leads, and if one follows closely, reason requires the rejection of religious belief. On both models, religious belief turns out to be irrational. In this piece, I summarize and examine these two interpretations, and using José Maia Neto's exposition of Bayle as an Academic skeptic, I propose a third that I call “common sense” skepticism. On this reading, reason does indeed use logic to form and sort beliefs – including some religious beliefs – but only holds to those beliefs fallibilistically. There may be a category of religious beliefs, however, that can only be recommended by faith; these would be religious beliefs that lack rational evidence, or exhibit rational contradictions. On this view, while some religious beliefs may turn out to be rational, there will likely be a set of core religious beliefs that come out to be irrational. I end by highlighting a suggestion by Bayle that reason itself may recommend revelation as a guide to belief, conferring a kind of intermediate rationality even on apparently irrational religious beliefs. (shrink)