We propose to develop, test, and prove the feasibility of a methodology for an inferential system for the generation of crash site likelihood maps. These maps will assist in the prioritization of candidate aircraft crash sites to be searched by Search and Rescue (SAR) operations. The crash site likelihood maps will be created by the fusion of the knowledge and experience of experts in aircraft crash site identification, together with, among others, knowledge about weather condi-tions, terrain information, and aircraft models. These maps will indicate the most likely areas where an aircraft may have crashed, and will allow SAR operations to focus their resources in these areas first, leading to faster and more efficient rescue operations. The proposed work di-rectly and innovatively addresses NASA's Search and Rescue (SAR) mission, by targeting the improvement of SAR operations. This proposed approach offers the innovation of the intelligent fusion of a variety of data, information and knowledge sources to generate crash site likelihood maps.