This thesis is a study in the theory of reasons. In the main, it comprises three chapters that investigate a single question from different angles, namely, the question of what it is that makes reasons possible. In Chapter 1, I propose a theory of what makes reasons possible. The theory progresses by defending a version of the thesis that ought implies can against an argument by John Gardner. What emerges from this defence is a formal model of possibility for facts that qualify as reasons for action. This contributes a new analytical test to the existing literature on normativity in respect of the question of what it is for a fact to be a reason. In Chapter 2, I take up the question of whether legal rules are possible reasons for action. They are, indeed, commonly regarded in this way. Yet are such rules reasons for action themselves (the reflexivity thesis) or is a statement of a rule a way of paraphrasing reasons that we may already have (the paraphrastic thesis)? I argue for a version of the paraphrastic thesis. In Chapter 3, I examine an influential methodology in the philosophy of law that underpins some influential theories on how legal rules can be possible reasons for action. I focus specifically on conceptual analysis as the methodology behind the requirement of compliance in Joseph Raz's theory of mandatory norms. This requirement holds that if such norms are to be reasons, then they must allow for the possibility that one can do as a reason requires one to do because it requires one to do it. I argue against this view, holding that the requirement is impossible on methodological grounds. What unifies the chapters of this thesis is a concern for what makes reasons possible. The take-home message is threefold. First, not all possible facts are possible reasons (Chapter 1). Second, legal rules are not themselves possible reasons (Chapter 2). And third, if we misuse the methods of conceptual analysis, we will end up with impossible reasons on our hands (Chapter 3). My tactic is to begin broadly by examining facts as possible reasons, then narrow down to one kind of reasons, namely, legal ones, and then step back to diagnose one source of impossible reasons.