We often refer to road fatalities without knowing exactly what injuries are responsible for them. Based on the Rhone Road Trauma Registry this paper sets out to describe the topography, nature and frequency of the fatal injuries sustained by car drivers. Mean annual mortality at the wheel of a car, computed by dividing the total number of drivers killed(n = 383) by the population of the Rhone Department (1.6 million) during the period 1996–2004 was5.41 males per 100,000 and 1.41 females per 100,000, with 78% of the casualties residing in the Department. A reduction has been observed since 2003.Three-quarters of the casualties died at the scene of the crash. The results confirm the effectiveness of seat belts. The observed lethality was 0.43% for unbelted drivers and 2.7% for belted drivers (RR = 0.16 [0.12; 0.21]).The injuries were analyzed for the 287 killed drivers whose deaths could be explained by the described injuries (at least one AIS 4+ injury). Of these, 41% had an ISS of 75 (at least one AIS 6 injury), 21%had an ISS of between 40 and 74, 33% an ISS of between 25 and 40, and 6% an ISS of between 16 and 24.In the case of all the AIS 4+ injuries, the three most frequent locations for injuries were the thorax only(30% of casualties), the head only (23%) and a combination of the two (18%). Abdominal injuries occurred in only 10% of casualties and spinal injuries in 9% of casualties. In the thorax, the most common injury was flail chest with haemothorax or pneumothorax. In the case of the head, the most frequent injuries were to the brain (haemorrhage, haematoma and axonal injuries). Complex fractures of the base of the skull were the second most common craniocephalic injuries. In spite of the use of restraint devices, the thorax and head are still the priority vital areas for protection in the case of car drivers. For one in four of the fatalities, death cannot be explained by any of the injuries we know about. As road traffic accidents are considered to be a cause of death in their own right, autopsies are rarely performed in France on such fatalities. This means we do not know about any injuries that cannot be detected by an external examination of the cadaver.