The relationship between crash severity and injury level is illustrated using the Restraint System Evaluation Study (RSES) data and Texas police-reported data. The RSES data are used to demonstrate that the probability of an injury (or fatality) is a function of both (a) the risk of injury, given a set of crash factors, and (b) crash exposure, or the chance of those factors occurring. The Texas data are used to demonstrate that crash exposure can change with time. Changes in crash exposure can alter the distribution of injuries (or fatalities) independently of motor vehicle standards or design. Texas and CPIR data are used to demonstrate that the risk of fatality is associated with many crash factors other than the traditional speed-related measures.