Frontal crashes and near-side crashes were compared and found to be significantly different events. In a frontal crash, the energy to be dissipated from the occupant is constant for a given speed. In a side crash, the energy transferred to a struck-side occupant depends highly on his interaction with the door. That difference has important implications on the choice of countermeasures, injury criteria, and subsystem tests. In a frontal crash, chest and abdominal injuries occur in the “second” impact when the occupant, acting like a free-flight mass, strikes the interior. Padding can absorb some of the free-flight energy, reduce the impact force, and provide earlier and longer contact of the occupant with the interior. The earlier contact decreases the differential velocity of the occupant to the interior, and the longer contact allows more time and greater distance to dissipate the kinetic energy. These improved energy management effects are reflected by lower deflection-based and acceleration-based chest injury criteria in a frontal crash test when padding is added. In a side impact, chest and abdominal injuries occur when the stationary occupant is “punched” by the encroaching door. Padding also provides an earlier and a longer contact of the occupant with the encroaching door, which prolongs the side impact punch. The prolonged punch can actually increase the net energy transferred from the door to the occupant, which may be reflected by higher deflection-based chest injury criteria. Despite the greater energy transfer, padding actually reduces acceleration-based injury criteria, such as TTI(d), in the same side impact crash test. Traditional subsystem tests, such as sled or pendulum tests, can mimic most effects on an occupant of a frontal crash, but they cannot mimic many essential characteristics of a side crash.