This dissertation explores the role of public opinion in U.S. presidential decisions to employ various alternatives in response to an international crisis. Presidents may choose from a range of force alternatives, including non-force alternatives, troop mobilizations, air strikes or ground assaults. Using the Poliheuristic Theory, I argue that public attitudes toward the use of force in a given crisis play a key role in the decision making process leading to such choices. The direction and intensity of public opinion is driven by a relative value assessment by which the public determines whether the benefits of a use of force are worth the costs. Presidents are aware of this relative value assessment and rule out crisis responses that are likely to violate the public's preferences in the first stage of the decision making process. In the second stage, presidents choose among the remaining alternatives by weighing the relative merits of each with respect to military and international-strategic implications. To test hypotheses following from this theoretical argument, I employ two methodological approaches. The first is statistical analysis. I develop a new data set of presidential crisis response choices and expand an existing data set on U.S. public attitudes toward the use of force, from 1949 to 2001. Using two extant data collections identifying international crises, I conduct Ordered Logit analyses, which produce results that are largely supportive of the hypotheses. The second methodological approach is the case study method. I conduct two detailed case studies of decisions to use force in Bosnia (1995) and Afghanistan (2001). These analyses are also supportive of the theoretical argument. I conclude that presidents are largely responsive to public opinion in the selection of crisis responses.