In this paper, I examine the presumption that Mill endorses a form of metaethical non-cognitivism. I argue that the evidence traditionally cited for this interpretation is not convincing, and suggest that we should instead remain open to a cognitivist reading. I begin, in Section I, by laying out the ‘received view’ of Mill on the status of practical norms, as given by Alan Ryan in the 1970s. There is, I claim in Sections II and III, no firm textual evidence for this reading of Mill: his remarks on ‘art’ and ‘science’ do not show the metaethical commitments they have been taken to. Neither is there firm textual evidence for a cognitivist reading. However, I suggest in Section IV, a non-cognitivist interpretation suffers from the fault of anachronism, and is difficult to reconcile with a clear commitment in Utilitarianism IV.3 to the possibility of evidence being given for the desirability of pleasure. A cognitivist reading would not suffer from these faults, and on that basis, I conclude that we should think further about what a cognitivist reading of Mill might amount to.