Because the break‐up of conglomerates typically produces substantial increases in shareholder wealth, many commentators have argued that the conglomerate form of organization is inefficient. This article reports the findings of a number of recent academic studies, including the authors' own, that examine the causes and consequences of corporate diversification.Although theoretical arguments suggest that corporate diversification can have benefits as well as costs, several studies have documented that diversified firms trade at a significant discount from their single‐segment peers. Estimates of this discount range from 10–15% of firm value, and are larger for “unrelated” diversification than for “related” diversification.If corporate diversification has generally been a value‐reducing managerial strategy, why do firms remain diversified? One possibility, which the authors label the “agency cost” hypothesis, is that top executives without substantial equity stakes may have incentives to maintain a diversification strategy even if doing so reduces shareholder wealth. But, as top managers' ownership stakes increase, they bear a greater fraction of the costs associated with value‐reducing policies and are therefore less likely to take actions that reduce shareholder wealth. Also, to the extent that outside blockholders monitor managerial behavior, the agency cost hypothesis predicts that diversification will be less prevalent in firms with large outside blockholders. Consistent with this argument, the authors find that companies in which managers own a significant fraction of the firm's shares, and in which blockholders own a large fraction of shares, are significantly less likely to be diversified.If agency problems lead managers to maintain value‐reducing diversification strategies, what is it that leads some of these same firms to refocus? The agency cost hypothesis predicts that managers will reduce diversification only if pressured to do so by internal or external mechanisms that reduce agency problems. Consistent with this argument, the authors find that decreases in diversification appear to be precipitated by market disciplinary forces such as block purchases, acquisition attempts, and management turnover.