Throughout his career as moral philosopher Hare has insisted that there is a rational way of arriving at substantive moral judgements. Hare develops this view - first presented in ' The language of morals' (1952) and ' Universalizability' (1955) - into the claim that rational agents are required to adopt utilitarian solutions to moral disputes. In ' Freedom and reason ' (1963) this claim is defended with reference to the view that the formal features of moral language (universalizability and prescriptivity)commit moral agents to a certain method of reasoning, and that this method of reasoning, when conjoined with facts about people's desires and preferences, leads us to accept substantive moral judgements consistent with those required by a form of utilitarianism. This view features throughout Hare's subsequent work, but the argument for it undergoes change. This means change in the defence of the claim that the meta-theory Universal Prescriptivism is consistent with a form of normative utilitarian theory, as this claim is argued for in 'Ethical theory and utilitarianism' (1976) and 'Moral Thinking' (1981). I shall endeavour to trace the chronological development of Hare's thinking, and will concentrate on developments in the argument for a theory of act-utilitarianism. I shall argue that the argument for utilitarianism gives rise to two major problems which arise from a specific feature of the argument, namely, the attempt to run the resolution of bi-lateral and multi-lateral cases of conflict along lines analogous to the resolution of conflict in the single-person case. Hare's argument requires that a decision-maker must identify the person with whom he reverses roles as himself, and that he must be prepared to concede that the things his recipient has good reasons for wanting are also reasons for him to want the same things. I argue that it is not possible to make coherent sense of the identity of the person in the reversed-role situation and that the motivational states a decision -maker is expected to deem 'his own' are not properly states of himself. If I am right, the 'identity'-question sits at the root of a motivational gap in Hare's theory.