This study examines public response to tornado warnings through an application of the framework laid out by Mileti and colleagues (2000). A qualitative approach was adopted to supplement our knowledge of public warning response with detailed descriptions of how individuals interpret and react to risk information. Following tornado outbreaks, data were collected from individuals in regions surrounding New Orleans, LA, and Springfield, MO, using in-depth interview guides. As each region is characterized by significant diversity, researchers developed a purposive sampling strategy to ensure the collection of representative data. Interviewees (n=40) were asked about how they received, interpreted, and responded to warning information. Researchers then used content analysis to analyze these data in order to evaluate and supplement Mileti's model. Ongoing analysis confirms Mileti's model, yet reveals a high degree of complexity with regard to a) stage transition, b) interaction and communication, and c) social factors. Findings from this study will contribute to the development of quantitative models intended to establish end-user policies. These policies will guide the deployment and use of radar technology currently under development by the Engineering Research Center for Collaborative Adaptive Sensing of the Atmosphere (CASA). From a forecasting perspective, findings suggest that a meteorological approach to public safety could be greatly augmented through the incorporation of social science methods and data. For instance, paying greater attention to how cultural myths about tornado threats shape risk communication could improve the effectiveness of watches and warnings.