The claim that we are subjects of experience, i.e. beings whose nature is intimately bound up with consciousness, is in many ways a plausible one. There is, however, more than one way of developing a metaphysical account of the nature of subjects. The view that subjects are essentially conscious has the unfortunate consequence that subjects cannot survive periods of unconsciousness. A more appealing alternative is to hold that subjects are beings with the capacity to be conscious, a capacity which need not always be exercised. But this view can itself be developed in more than one way. The option I defend here is that subjects are nothing more than capacities for consciousness, a view I call the ‘C-theory’. Although the C-theory supplies us with a potentially appealing account of the nature of subjects (and hence ourselves), there are challenges to be overcome. Olson has argued that identifying ourselves with what are, in effect, parts of human organisms leads to a variety of intolerable problems. I suggest that these problems are by no means insuperable. Bayne and Johnston have argued that identifying subjects with experience-producing systems is confronted with a different difficulty. What if these systems can produce multiple streams of consciousness at once. Whatever else they may be, aren't subjects the kind of thing that can have just one stream of consciousness at a time? In response I argue that this is true in one sense, but not in another. Once this is appreciated, the notion that a subject could have several streams of consciousness at once no longer seems absurd, or impossible.