This paper revisits one of the key ideas developed in The Structure of Scientific Revolutions. In particular, it explores the methodological form of incommensurability which may be found in the original edition of Structure. It is argued that such methodological incommensurability leads to a form of epistemic relativism. In later work, Kuhn moved away from the original idea of methodological incommensurability with his idea of a set of epistemic values that provides a basis for rational theory choice, but do not constitute an algorithm for such choice. The paper also explores the sceptical basis for the epistemic relativism of the original view that Kuhn proposes in Structure. It suggests that the main sceptical rationale for such relativism may be avoided by a particularist and naturalist conception of epistemic normativity. When this approach is combined with the appeal to external methodological standards endorsed by the later Kuhn and his critics, the epistemic relativism of Structure may be completely repudiated.