Historically, the most important rationale for coercive public health measures has been the prevention of disease and injury to others. However, as noncommunicable diseases and accidents have assumed increased importance as causes of morbidity and mortality, and as the connection between noncommunicable diseases and accidents and individual practices such as smoking and drinking has become more apparent, a new line of argument based on social costs has emerged. My purpose is both to describe and evaluate the social-costs argument, to explain why it has become so popular, and to show what must be done to make it consistent with its own utilitarian criterion.