The idea that a receptor can produce signalling without agonist intervention and that several antagonists can be 'active' in repressing such spontaneous activity is contained in the concept of ligand-induced conformational changes. Yet, this idea was neglected by pharmacologists for many years. In this article, we review the events that brought inverse agonism and constitutive activity to general attention and made this phenomenon a topic of current research. We also suggest a classification of antagonists based on the cooperativity that links their primary site of interaction with other functional domains of the receptor.