Australasian Journal of Philosophy. forthcomingExpected value maximization is something like the gold standard in decision theory. For general theories of normativity, the notion of a reason (for action) is much more central. Surprisingly little work has been done on how reasons for action relate to expected value maximization. Can one have most reason to do something that would not maximize expected value? Or is there a way of securing some kind of correspondence between expected value and what one’s reasons support? In this paper, I develop a framework for giving a positive answer to the latter question. More precisely, I defend the following thesis: the totality of one’s possessed reasons always lends strongest support to an action with maximal expected value— provided one’s value function adequately reflects the strength of one’s reasons.