The first part of this investigation is concerned with the results of the autopsies of 70 passengers. The material has been gathered from head-on collisions involving passenger cars of various types, where no safety belts were used.The injury mechanisms are divided into six typical groups according to the parts of the vehicles struck by the passengers.It appeared that the most important factors deciding the type of injury mechanism were the construction of the engine hood, the windshield and frame, the instrument panel and the seat, not considering intruding objects and parts of other vehicles which may cause injuries.The second part of this investigation has as its objective the corroboration of these observations in sled tests of head-on collisions using dummies; the results obtained are measured as forces acting in the head, the chest, and the thighs. In addition, the movement of the occupants is studied when the design of the instrument panel is changed.FOR DECADES, automobile manufacturers and scientific institutes have been making efforts to reduce the injury risk of car occupants. Besides statistical surveys of the frequency and causes of injuries, numerous theoretical projects have been carried out with mathematical and mechanical models, as well as experimental investigations with both laboratory devices and full-scale crash test facilities. However, many of these research projects suffer from the unsatisfactory simulation of the most important part in the system, namely the human being. It is now recognized that crucial problems can be illuminated by exact study of the actual road traffic accidents and of injured car occupants-in other words, by field accident investigation.In a previous publication (1)*, the various mechanisms were described in which the leading injuries to drivers of cars involved in head-on collisions were caused. The reconstruction of fatal accidents made it clear that there were five principal injury mechanisms, and showed that a decisive influence on the type of injury was exercised by the design and position of the steering, the instrument panel, and the seat.The following pages will be a summary of the findings of the investigations of 70 accidents leading to the death of the front-seat passenger in vehicles colliding head-on. The autopsy findings were correlated, as before, with the traces left on the vehicles. In many cases a simple dummy was used for the reconstruction of the “second collision.”The following vehicle parts are decisive factors determining the type of injury sustained by front-seat passengers: windshield, windshield frame, front hood, other vehicles (primarily truck frames), instrument panel, and seat.The deduction of typical injury mechanisms from these factors proves to be very difficult. In the driver's case, the first four components play a subordinate part, being virtually replaced by the steering assembly. In the case of the front-seat passenger, the larger number of effective components and consequent diversity of possible patterns of movement compels greater caution in making a schematic division into typical injury mechanisms. Nonetheless, the attempt to do so will now be made, and it is hoped that this will form a basis for discussion.The injury mechanisms deduced in this way are the basis for experimental and theoretical investigations, the object of which is to reduce the injury risk to occupants. First of all, therefore, impact experiments were made with the impact sled and dummies, in order to produce uniform and if possible, reproducible conditions which would give an insight into the movements made by and the forces acting on the dummy, as they occur on impact with the various components contacted by the occupants themselves. The object of this is to find the technical prerequisites for improved impact protection for the occupant in head-on collisions.One particular difficulty is that it is impossible to produce injuries to the skeleton and soft parts of dummies at present available, which makes experiments with cadavers seem necessary. A report of this is made elsewhere (2).Even when great care is taken to obtain uniform test conditions, experiments with dummies produce a relatively wide spread of measurement values (3). Today it is not yet possible to obtain a reliable correlation of the values with the injuries mentioned above, despite considerable advances in biodynamics and the construction of the dummies.The study of the movements of the dummy proved to be particularly valuable. It became apparent that the movements of the unrestrained dummy were basically determined by the instrument panel, and to a certain extent, the seat.Thus, the second part of this study is chiefly devoted to the latest findings relating to the design of the instrument panel.