It is generally assumed that whatever R. M. Hare might be he is not a naturalist or a descriptivist. This means, among other things, that he believes that action-guiding statements ("ought" statements) cannot be derived from factual statements ("is" statements) and that moral predicates, such as "is good" cannot be defined in terms of any nonmoral predicates. Much of Hare's writing on metaethics is an explicit defense of nondescriptivism. Some of his recent work, therefore, comes as a surprise, for it appears to be naturalistic. Hare's central concern has been to develop a theory which is a form of rational nondescriptivism.1 The system which is supposed to meet this objective is universal prescriptivism. Roughly, the universalizability provides the rationality and prescriptivism excludes descriptivism. It is one thesis of this paper that these two aspects of Hare's program have become inconsistent as a result of later developments. Hare's determination to use the principle of universalizability to generate utilitarianism has led to the destruction of the antidescriptivist and antinaturalist element in prescriptivism. In Section VI, I shall argue that Hare's concept of practical reasoning also leads to naturalism.