9410683 Songer The objective is to develop an automated prediction tool for public sector agencies to select appropriate projects for design- build facility acquisition. At the core of the tool is a analytical model which predicts the probability of success for projects selected for design-build. The model will be built using two techniques, multiattribute analysis and logistic regression. The model will be built using case studies from previous design-build projects of Building Research Board members. Rigorous testing of the model and tool will be performed by comparing predicted to actual project performance. Design-build acquisition methods are relatively new to public sector agencies. Providing a tool to select appropriate projects for design-build will significantly enhance current selection processes. Appropriate project selection will greatly improve public sector competitiveness in facility acquisition. ***