Executive Summary The first workshop of the project on Prediction in the Earth Sciences: Use and Misuse in Policy Making was held July 10-12 at the National Center for Atmospheric Research in Boulder, Colorado. The purpose of the workshop was to present, discuss, and synthesize findings from nine case studies from the earth sciences on the role of prediction in the policy process. The overarching goal of the project is to improve policy and decision making by improving the way predictions are used in policy processes. We plan to do this by developing guidelines for the beneficial use of scientific prediction in the policy process, especially in the area of environmental policy problems. We seek to develop a set of criteria and an analytical framework to help policy makers assess the actual and potential use and value of scientific prediction for different policy problems. The criteria and framework are intended as tools to allow policy makers to be better consumers of predictions, and to provide feedback from policy makers to researchers. The workshop was organized to bring together people who have expertise in both the scientific and policy aspects of relevant case histories, plus several discussants with general insight into the issue. The goal of the meeting was aptly reinforced by one participant who noted the contrast between the depth to which policy makers rely on predictions versus the lack of depth in society's understanding of the consequences of that reliance. Each case history presenter was asked to prepare a draft paper addressing the following questions for their particular case. The questions were introduced as a means to focus initial discussions and to guide preparation of the case histories, with the expectation that through discussion of the cases, the initial questions would likely be revised or replaced in order to guide future work. What societal problem was the predictive research intended to address? What were the original scientific justifications for the research? What were the original policy justifications? How were these justifications incorporated into the research program as goals? How was (is) the research program linked to the information needs of policy makers, during both program planning and program implementation? Was the research program modified at any time during its conduct in response to evolving scientific and policy needs? How successful was (is) the research in meeting scientific goals? How successful was (is) it in meeting policy goals? Upon what criteria is this evaluation based? How can this degree of success or failure best be explained? What alternatives to prediction and/or predictive modeling are available for providing guidance in meeting the needs of policy makers and other potential users of predictive information? The draft case histories were completed before the workshop and circulated in advance to all workshop participants. The workshop itself focused on presentation and discussion of each case. The group worked together to identify important policy issues and questions common to the cases. A common frame of reference across the cases is a first step in the development of practical guidance on the use and miseuse of prediction. This report contains summaries of the case histories presented at the workshop as well as the common principles distilled from the cases by the workshop participants. The case studies themselves will be revised and presented in a second workshop, tentatively scheduled for mid- to late 1998. We especially thank D. Jan Stewart, rapporteur, and Miles Mercer, student assistant, whose efforts contributed in significant ways to the workshop's successful organization and outcome. We also thank Baat Enosh for expert work related to publishing this report on the World Wide Web. The workshop was sponsored by the Geological Society of America and the Environmental and Societal Impacts Group at the National Center for Atmospheric Research.