In updating the regional transportation plan for the Eugene-Springfield area in Oregon, the Lane Council of Governments (LCOG) developed a series of six alternative plan scenarios to test the effectiveness of various strategies (transportation demand management, land use, system changes). To measure the effectiveness of scenarios performance criteria were developed. While safety was a broad issue of interest to policy makers and the public, no safety criterion was found that could be modeled and forecasted in a practical manner. In 2001, a research project was undertaken by the University of Tennessee to develop and test practical tools for assessing safety impacts of transportation plans for urban areas. As part of this research, LCOG was engaged to apply a set of accident rates to its alternative long-range plan scenarios. This study concludes that, in general, at the point where issues raised as part of this application are addressed through future research, the approach of using a standard travel-demand forecasting model and a series of crash forecasting spreadsheets could provide a practical means for assessing the safety impacts of long-range plans in small and medium-sized communities. The application effort yielded several specific observations relevant to future efforts. Key issues include lack of transferability of crash data, availability of and easy access to detailed crash data for a given urban area, refinement of continuous functions for developing crash forecasts, and refinement of intersection crash analysis methods.