While epistemologists routinely employ disbelief talk, it is not clear that they really mean it, given that they often equate disbelieving p with believing ¬p. I argue that this is a mistake-disbelief is a doxastic attitude of rejection and is distinct from belief (and withholding). I first clarify this claim and its opposition, then show that we must distinguish disbelieving p from believing ¬p in order to account for the fact that we continue to hold doxastic attitudes toward propositions that we reject. After defending this argument against some possible objections, I examine several cases that reveal disbelieving p to be not only non-identical to believing ¬p, but independent of that attitude as well. Finally, I sketch some immediate and potential consequences of recognizing disbelief as a distinct doxastic attitude, particularly for work on epistemic rationality.