Current vehicle and restraint designs consider a relatively small percentage of the variation that exists within the population of humans and the distribution of crashes. The breadth of these variations, coupled with the added complexity of ever-changing societal trends, will not permit designs developed using conventional techniques to realize established goals for injury and fatality reduction. Within the realm of passive and integrated safety, widespread implementation of modeling into the design and evaluation process provides the only viable means of achieving significant reductions in injury and fatality. Ultimately, the versatility and efficiency of computational modeling will establish a new paradigm in which injury is described by causal mechanisms including statistical representations of population variations. While this transformational change will be guided and directed by simulation technology, it will require commensurate evolution of techniques within the fields of experimental biomechanics, crash reconstruction, and epidemiology.