In recounting his early research agenda, Skinner (1956) described the serendipitous event that led to the discovery of operant extinction curves. The pellet dispenser on his apparatus jammed, leaving lever pressing unreinforced. The cumulative record obtained under this defective procedure was nonetheless orderly. Lever pressing did not stop immediately when reinforcement was withdrawn. Rather, it continued for a while, decreasing as exposure to the new contingency proceeded. As Skinner noted, the record was one of “pure behavior”(p. 226), uninterrupted by food delivery or consumption. He recalled, too, his excitement on observing this first extinction curve, suggesting to him that he “had made contact with Pavlov at last!”(p. 226). Skinner’s serendipitous finding not only made contact with Pavlov, but it turned out to be one of the most reliable effects in the study of learned behavior. Furthermore, the effect has considerable generality. Extinction is found across classical and operant conditioning, in basic and applied research, and in practice. It also is widespread in the animal kingdom, having been reported in organisms ranging from invertebrates (eg, Abramson, Armstrong, Feinman, & Feinman, 1988) to Homo sapiens. The effects of extinction, however, are neither unidimensional nor simple. Eliminating reinforcement can diminish not only the previously reinforced response, but also others that are related to it. Although responding is eliminated in the sense that it is no longer observed, the extinguished response is quickly reestablished under appropriate circumstances, leading some to describe extinction in terms of discriminative stimulus control and others to consider it as evidence for a learning–performance distinction (eg, Hull, 1943). Skinner’s subsequent research on extinction (eg, 1933b) revealed something of its synergistic relation to other contingencies of reinforcement. When he alternated periods of reinforcement and nonreinforcement, the result was the familiar scalloped pattern of responding, and the fixed-interval (FI) schedule was born—a direct descendant of operant extinction. Its sibling, the variable-interval (VI) schedule, was created by substituting variable for fixed interreinforcer intervals (IRIs). Other analyses of the synergistic relation between extinction and other events in the organism’s proximal and distal histories have elaborated the generative effects of extinction. This chapter is a review of research from basic and applied behavior analysis bearing on both the eliminative and generative effects of the extinction of operant behavior.(PsycInfo Database Record (c) 2022 APA, all rights reserved)