Considerable effort has been expended particularly over the previous decade in the area of examining pre-crash factors in an attempt at developing appropriate highway safety countermeasures. This report summarizes the review of this field by a select panel of highway safety researchers and includes recommendations for NHTSA-sponsored research in accident causation over the next several years. After attempting to clarify the significance of the “cause” of an accident, the report outlines various approaches to accident causation research and indicates the relevant data needs. The extent and type of exposure information required is addressed. A detailed critique of the major research efforts in this field is provided. This critique centers mainly on the work done by Indiana University, Cornell Aeronautical Laboratory, Inc., Operations Research, Inc., and the University of Miami. Finally, short and long-range recommendations for NHTSA-sponsored research are presented. These include in general utilizing the NASS system to the fullest extent possible. Several specific studies recommended include: examination of data needs; survey of the possible existence of this required data; exploration of various alternative approaches to data collection (e.g. simulation); establishing appropriate quantifiers (e.g., what defines “following too closely”); determining exposure requirements; upgrading the statistical techniques utilized in this field; ascertaining the effectiveness of various countermeasures; and examining alternative information retrieval systems. In addition, several small open-ended research contracts for just plain “thinking about” accident causation methodology were recommended.