The goal of this research is to understand the conditions within organizations that lead to crises so that preemptive actions can be taken to avert them. Although industrial crises (e.g., Bhopal, Three Mile Island) have received increasing attention over the past decade, much of the evidence collected to date is based on anecdotal accounts. As a result, there is no overall theoretical framework in which to place the present understanding of organizational crises, and little that can be said in terms of empirically-based recommendations for promoting behaviors designed to avoid the next catastrophe. The proposed research plan involves a longitudinal study that will compare the decision making processes of four organizations in the chemical and electric power industries. Two of these organizations will have experienced major crises and exhibit poor safety records, whereas the other two will have much better safety records and no experience of a major crisis. An in-depth analysis of organizational strategies and decision processes that have affected the performance of these companies over a ten-year period, based on both available secondary data sources and personal interviews, is proposed; variables to be considered include the degree of competitiveness within an industry, the rate of diversification within a firm, and the degree to which decision-making authority is centralized. The researchers emphasize that the proposed work constitutes a pilot study, designed to test research methods while developing an initial data base. However, the results should provide important theoretical insights into the nature of industrial crises and serve to establish preliminary guidelines for further studies in the field.