In the Critique of Pure Reason, following his elucidation of the ‘postulates’ of possibility, actuality, and necessity, Kant makes a series of puzzling remarks. He seems to deny the somewhat metaphysically intuitive contention that the extension of possibility is greater than that of actuality, which, in turn, is greater than that of necessity (A230/B283). Further, he states that the actual adds nothing to the possible (B284). This leads to the view, fairly common in the literature, that Kant holds that all modal categories, in their empirical applications, are coextensive. I diverge from the common view. First, Kant is not committed to the coextensiveness thesis, understood as above. Instead, he espouses a weaker, epistemological version of the coextensiveness thesis, namely that what we can assert to be really possible is coextensive with what we cognize to be actual. Second, Kant's remarks are not intended to introduce a positive ontological thesis about the extensions of modal categories. Rather, he means to criticize a certain conception of modalities that was prevalent among his rationalist predecessors, i.e., the conception of modalities as various determinations that enter the intensions of concepts of things.