Understanding injury severity patterns in roadway crashes is important not only from the view point of treating crash victims, but also for directing the crash avoidance efforts of traffic safety agencies and motor vehicle manufacturers. The factor that is most discussed in this context is vehicle incompatibility. However, there are other vehicle-, occupant-, and roadway-related factors, too, that play roles in injury severity.

In order to investigate these factors in relation to the injury severity, this paper considers a two-vehicle crash as a ‘system’ with its elements:​ vehicles, drivers, and roadway. Some of the possible inputs (contributing factors) to this system are considered with a focus on injury severity of the driver as an outcome. The differences in weights, heights, and shapes, etc. of the crash-involved vehicles, vehicle speed, drivers’ ages and genders are the factors in question. Data mining the crash databases compiled by the National Highway Traffic Safety Administration (NHTSA) makes many important revelations. The association between the subject variables and driver injury severity is studied through contingency analysis. Configuration frequency analysis helps to identify patterns of injury severity. The main objective of the study is achieved by building a logit model that can be used to predict the likelihood of injury severity from a given set of vehicle-, driver-, and roadway-related crash characteristics.