A number of recent international situations have raised again questions regarding the usefulness of economic sanctions as an instrument of foreign policy. Sanctions continue to be applied in a variety of contexts, yet we have not developed a sufficient understanding of the processes involved to determine when, or even if, sanctions can "work." While a great deal has been written on the subject, there have been neither attempts to subject the theoretical arguments to empirical testing nor efforts to provide systematic theoretical explanations for the empirical results that have been produced. In this article, we attempt to address this shortcoming in the literature. We propose a theory of sanctions effectiveness that is based on the spatial model of bargaining in international crises and use this theory to derive a number of hypotheses regarding when sanctions should produce favorable policy outcomes. We then subject some of the derived hypotheses to an empirical test based on a large number of international disputes. The model suggests that while sanctions will not work in many cases, they can have a slight effect on the distribution of expected outcomes if the costs of the sanctions are sufficiently high relative to the values at stake. The available evidence appears to support these expectations.