At the outset of The Possibility of Altruism Thomas Nagel charts two paths out of the fundamental dilemma confronting metaethics. The first path rejects the claim that a persuasive account of the motivational backing of ethical judgments must involve an agent’s desires. But it is the second path, a path that Nagel charts but does not himself take, that is the focus of this essay. This path retains the standard account, upon which all motivation involves desire, but denies that desires are given prior to reason. Instead, these attitudes that motivate are themselves open to rational assessment. One reason for this focus is that many philosophers, including Quinn, Raz, and Scanlon, have come to reject the claim Nagel takes to block this path – that desires are somehow given prior to reason, hence are not in the relevant way proper objects of rational assessment. A second reason is that unlike the first path, this second does not require the rejection of the belief-desire theory, only the rejection of one assumption about the nature of conative attitudes. Unlike Nagel’s chosen path, then, the second holds out the prospect of reconciling ethical objectivity, internalism, and the belief-desire theory within a unified account. I argue that the account of desire found in Quinn, Raz, and Scanlon, augmented by aspects of Davidson’s account of propositional attitudes, yields a coherent account of the involvement of reason even in basic desires, an account that is well suited to Nagel’s intriguing path not taken.