The distinction between rationalists and empiricists has become something of a punch-bag in recent decades, and we have been encouraged to pursue supposedly subtler analyses of purposes, influences and allegiances. One line of attack has taken the form of various attempts to divorce the “British Empiricist” Berkeley from either Locke or Hume, or from both Locke and Hume, and to hand him over to the Cartesians. The rationalist-empiricist dichotomy is then dismissed as a construct of nineteenth-century Kantians. My present purpose is to argue, on the contrary, that it is a distinction not only with ancient origins, but with a peculiar relevance to the interpretation of early-modern philosophy, not least that of Berkeley, just because of those origins. On the other hand it needs to be explained how such disagreement over the classification of Berkeley can have arisen.