Abstract: Lockups are an increasingly important element of M&A deals in the United States. We present, for the first time, descriptive data on lockup incidence, trends, and their relationship with Delaware case law. Prior commentators have used theoretical models to argue that lockups should have little or no impact on allocational efficiency in the market for corporate control. We use data from twelve years of M&A activity in the United States to show that prior models have little predictive power in real-world transactions. We then offer a new theoretical model of lockups that includes six "buy-side" distortions: agency costs, tax effects, informational effects, switching costs, reputational effects, and endowment effects for bidders. The implications of this new model suggest that courts and corporate boards should scrutinize lockups more closely than prior commentators have advocated.