From a biomechanical point of view, test criteria in current safety standards for passenger protection do not insure a sufficient over-all protection of quality.First, the deficiencies of data and criteria, responsible for biomechanical problems, are being analysed. Secondly, additional criteria are being defined, which we think are significant for a better over-all evaluation of restraint devices considering biomechanical facts. As a result of a dummy-crash-series, an analysis is presented, demonstrating correlations between the new defined and former criteria.The final aim is to develop a complete system of criteria. By the use of simple evaluation methods it would guarantee clear results concerning biomechanical properties of passenger protection systems. With this technique one can have correct biomechanical evaluations of restraint systems, gained from dynamic tests even with anthropometric dummies.