This paper discusses appropriate methodology for measuring the effect of an event on the value of a firm's equity. Th e key points are (1) cumulative abnormal returns do not measure the effect of an event on firm value if there are dividends during the event window; (2) it is generally appropriate to use pre-event parameters of the return-generating process even if the event alters the parameters during the event window, and (3) controlling for fact ors other than the return on the market portfolio improves the power of the estimation. The formula for the effect of an event on the value of a firm when there are dividends during the event window is developed a nd applied to a study of the effect of the Bhopal disaster on the value of Union Carbide. Copyright 1992 by MIT Press.