Classical methods for the estimation of antagonist affinity constants were developed under the assumption of one unique state for the receptor. The finding of receptor constitutive activity, which implies that at least two (one active and the other inactive) receptor states coexist at equilibrium, extended the concept of antagonism by distinguishing between neutral antagonists and inverse agonists. To account for the complexity introduced in the concept of antagonism, classical Schild and Cheng-Prusoff methods have been revisited within the two-state model of agonism. The resulting equations match the classical expressions for neutral antagonists but not for inverse agonists. It is suggested a revision of current routine procedures for antagonist affinity estimation.