This article examines Gibson’s concept of perceptual system and Reed’s concept of action system. After discussing several assumptions underlying these concepts, the ontological status of these systems is considered. It is argued that perceptual systems and action systems should be conceptualized neither as parts of an animal’s body nor as softly (temporarily) assembled devices; rather, they are best understood as animals’ abilities to achieve functional relationships, that is, as dispositional properties. This conceptualization entails that these systems are relatively permanent properties of the animal that are causally supported by, though not identical to, anatomical substrates. Further, it entails that it is the animal that perceives and acts, not its perceptual and action systems.