In a well-known paper, John Doris argues that situationist psychology impugns Aristotelian virtue ethics, which presupposes the existence of stable, situation-independent virtues. Maria Merritt responds that a Humean conception of virtue, which is situation-dependent, is immune to this criticism. However, she does not directly address two of Doris’ more trenchant objections, which he presents in the form of a dilemma. In this paper, I respond to Doris’ dilemma, using recent research in psychology and cognitive science to show that virtue ethics, as a dispositionalist, non-codifiable theory, represents a more empirically adequate moral psychology than the leading alternatives.