Injuries sustained by car occupants were studied on the basis of representative material comprising 29,000 accidents.As a unit of reference in defining typical accident categories, the relative collision speed (RCS) was defined, rendering it possible to take into account of real-life factors. The importance of the RCS as a categorizing method is discussed in relation to the equivalent test speed (ETS) in crash tests. The frequency of actual accidents with regard to five typical accident categories and the resulting passenger injuries are indicated.The influence exercised by the mass of the vehicle, its deformation characteristics, its turning away movement in collision, and its interior safety are compared in accidents involving four typical vehicle categories. The confrontation with an equivalent number of accidents whereby the passengers were belted clearly showed the benefit of safety belt usage.The collision speeds of 94 frontal accidents with fatal injuries to occupants are discussed. The paper specifies up to what speed safety tests for typical accident categories are to be carried out in order to include 90% of all accidents.