The last 3 decades of vigorous enforcement of U.S. antimerger laws is based on the market concentration doctrine, one of the oldest and most controversial propositions in industrial economics. This doctrine, which is an implication of oligopoly models in the tradition of Cournot ([1838] 1927) and Nash (1950), holds that the level of industry concentration is a reliable index of the industry's market power. The empirical implication is that a relatively high level of industry concentration, which in the presence of entry barriers is believed to facilitate intraindustry collusion or dominant-firm pricing, should be associated with relatively large industrywide monopoly rents.' Following Bain (1951), numerous The market concentration doctrine predicts that a horizontal merger is more likely to have collusive, anticompetitive effects the greater the merger-induced change in industry concentration. Since a collusive, anticompetitive merger generates an increase in the industry's quality-adjusted product price (or a decrease in factor prices), it also follows from the doctrine that the mergerinduced expected benefits to the product market rivals of the merging firms should be an increasing function of the concentration change. The empirical results of this paper, which are based on the industry wealth effect of a large sample of horizontal mergers, including cases found in violation of antimonopoly laws, fail to support this prediction. This conclusion is robust with respect to assumptions concerning the probability that a proposed merger will be prevented by the law enforcement agencies, and it continues to hold after transforming the industry wealth effect into a hypothetical, constant expected change in the industry's product price. The results imply that the levels of concentration and market shares found in the Department of Justice's merger guidelines are unlikely to identify truly anticompetitive mergers. * I wish to thank Paul Asquith, Gregg Jarrell, Rex Thompson, Walter Vandaele, the participants of the economics workshop at the Antitrust Division of the U.S. Department of Justice, and a referee of this Journal for helpful comments. Earlier drafts were presented at the 1983 meetings of the European Association for Research in Industrial Economics and at the 1984 meetings of the Western Finance Association. Financial support from the Social Sciences and Humanities Research Council of Canada (grant no. 494-83-0091) and the U.S. Federal Trade Commission is also gratefully acknowledged. 1. A closely related but somewhat less general prediction is that high levels of concentration should be associated with relatively high, supracompetitive product prices. Although