This paper is based on accident statistics, theoretical considerations and physical tests of safety bumpers and their components. The statistical analysis shows the typical bus frontal collisions, their frequency and the possible advantage of the safety bumper in the typical collisions. The theoretical considerations try to outline the possible requirements of a safety bumper:​ deformation capability, energy absorption capability, strength requirements relating to the bus structure behind the safety bumper system, compatibility requirements, etc. When specifying these requirements all bus categories, all kind of buses (low floor and high decker, small and large, etc.) should be considered together with their special features. There are welldefined theoretical connections between the length of the deformation, the energy absorption of the bumper and the average deceleration of the bus having safety bumper in a frontal collision. This deceleration is an important figure when regulating safety belts and seat strength in buses. The tests, the results of which are discussed in the paper include:​ pendulum impact tests of components of safety bumper systems, static loading tests of these components and full scale frontal impact tests with complete buses against concrete wall. The differences between the results of static and dynamic tests – carried out on the same components – are shown and discussed. It is emphasized that the bumper cannot solve all the safety problems belonging to frontal collision of buses, but it may be a useful, effective tool in some cases (avoiding underrun type accidents, reducing the decelerations below a certain impact speed, etc.)