Most of the thoracic injury criteria proposed these past years are based mainly on the analysis of data obtained from experiments with human cadavers and are related principally to rib fractures. And yet, the actual threat to life incurred in real accidents results from lesions to intrathoracic viscerae that are not necessarily correlated with the number of rib fractures. Moreover, the injuries sustained by cadavers during experiments are mainly rib fractures. In these cases, visceral lesions are unusual and not necessarily identical with those sustained by real accident victims.Preliminary results of measurements concerning rib strength and dynamic behavior of the cadaver thorax in frontal impacts show the rib cage offers a low resistance against compression and the vertebral column acts as a stop. The combination of these various data leads to questioning the validity of thoracic injury criteria based on the number of rib fractures observed on human cadavers.