Results of a one year field study initiated by the Swiss Federal Police Department in connection with a mandatory seat belt law effective January 1, 1976 are presented. 304 accidents with 153 killed and 257 severely injured (OAIS ≧ 2) belted occupants were analysed during this program. As belt effectiveness has previously been demonstrated in field investigations the following specific problems are discussed: Frequency and statistical significance of cases in which, with a high degree of probability, the belt had an adverse effect Relevant injury patterns Frequency and causes of belt failures including releasing difficulties after a crash Possible improvements. Furthermore, the applicability of Calspan's SMAC method is demonstrated and discussed. In case of car to car collisions the mass ratio is shown to be of importance. A correlation analysis between the velocity change during the collision phase (Δv) and the injury severity is performed. Special attention is given to frontal crashes with no or minimal passenger compartment intrusion. Severe injuries in such cases can usually be attributed to one or more of the following reasons: submarining, excessive slack, inadequate mechanical compatibility between belt system compliance and mechanical properties of the chest (e.g., old age), overloading due to backseat passenger.