ignette and autobiographical recall studies have often been used to test models of the appraisals associated with specific emotions. Recently, critiques of both methodologies have called into question the applicability of appraisal theory to naturally-occurring emotional responses. This study examined supporter's responses to Ross Perot's withdrawal from the 1992 presidential race to assess the extent to which appraisal models accurately capture responses to a naturally-occurring event. Supporters in Riverside County, California (N = 227) completed questionnaires concerning their interpretations of the (...) event, their initial emotional reactions, changes in their reactions in the weeks immediately following Perot's withdrawal, and their plans. Appraisal models correctly predicted the dimensions of agency and coping potential that served to differentiate sadness from anger, and the appraisals and plans associated with hope. Appraisal models did not predict the strong focus of sad supporters on their past goals and beliefs, and the plans associated with sadness and anger. Changes in emotional intensity occurred as a function of changes in goals and plans rather than merely as a function of time. These findings demonstrate the applicability of appraisal theory to naturally-occurring emotional responses, but also show that models may understate the complexity of the planning processes associated with emotions. (shrink)