Perhaps death’s badness is an illusion. Epicureans think so and argue that agents cannot be harmed by death when they’re alive (because death hasn’t happened yet) nor when they’re dead (because they do not exist by the time death comes). I argue that each version of Epicureanism faces a fatal dilemma: it is either committed to a demonstrably false view about the relationship between self-regarding reasons and well-being or it is involved in a merely verbal dispute with deprivationism. I first provide principled reason to think that any viable view about the badness of death must allow that agents have self-regarding reason to avoid (or seek) death if doing so would increase their total well-being. I then show that Epicurean views which do not preserve this link are subject to reductio arguments and so should be rejected. After that, I show that the Epicurean views which accommodate this desideratum are involved in a merely verbal dispute with deprivationism.