Market definition remains a critical part of most antitrust cases, despite attempts to obviate it. While many commentators have argued that a particular policy change may theoretically broaden or narrow antitrust markets, and thereby affect substantive antitrust enforcement, none has empirically tested the hypothesis. This article does so—albeit crudely—for one class of antitrust cases, mergers. It finds that, in practice and on average, product markets in Clayton Act cases have narrowed since the 1960s and 1970s. This narrowing is not necessarily desirable or undesirable, but it is real. Causes and implications of this narrowing are then explored, including its effect on legal rules like the structural presumption and the rule against out-of-market efficiencies. Two examples from the banking industry illustrate how narrowing can affect substantive antitrust outcomes, with the Department of Justice’s narrower market definition finding greater competitive harm, counting fewer efficiencies, and requiring more divestitures than the Federal Reserve Bank’s broader markets. This divergence suggests that Clayton Act rules are applied in a systematically different way today. As a result, it may be more appropriate to apply these rules—especially the rule against out-of-market efficiencies—to product markets akin in size to those in Philadelphia National Bank and other foundational cases.