Functionalist identity theorists (functionalists, for short) argue that if physical states of the central nervous system (CNS) have the same function (the same input and output) as pain, pains should be identified with those physical states. Many objections have been raised against this position. My aim in this paper is to defend it against opponents who argue that it leads to an absurd result: the ascription of pains to things which cannot reasonably be thought to be capable of suffering, or of having any conscious states. In doing this, I will outline a version of the functionalist position which I think is plausible. It departs in several respects from versions commonly found in the literature.