According to a National Highway Traffic Safety Administration (NHTSA) study, small overlap and oblique frontal crashes account for 25% of fatal accidents in the United States. This is a large proportion, and comprises the accident mode with the highest fatality rate. The fact that occupants collide with the cabin is a factor in this. The purpose of the present research is to recreate the injuries that occur to occupants in actual accidents involving small overlap and oblique frontal crashes, and to formulate a method for evaluating them. By enhancing vehicle body performance and occupant restraint performance, the goal is to reduce fatalities and injuries. In order to verify the circumstances of injury during small overlap and oblique frontal crashes, crash simulations were performed by numerical computation. A human body model was also placed in the driver's seat to verify the occupant movement and extent of injury. The oblique mode that the NHTSA is slated to adopt was first verified using numerical simulation, but collision on the occupant's chest could not be confirmed. For the present research, therefore, verification of the angle and speed parameters was carried out so that occupant movement could be recreated in a way that conformed to actual accident circumstances. The result of simulation was that occupant movement with respect to the vehicle together with vehicle body deformation showed the occurrence of collision with the occupant's chest when the evaluation vehicle (a passenger car) was impacted by an OMDB with an overlap (LAP) of 25%, at an angle of 30°, and at a speed of 110 km/h. The collision occurred in the same place, with injury to the same areas, and with the same degree of injury as shown under actual accident circumstances in a small overlap and oblique frontal crash. In light of these results, it was confirmed that injuries would be reduced by increasing the body strength of the evaluation vehicle to reduce deformation and by installing air bags on doors where oblique loading affected occupant movement. Numerical simulation of the above modes was performed using a passenger car that received the highest evaluation from the Insurance Institute for Highway Safety (IIHS). While vehicle body deformation was evaluated at the GOOD level even in oblique mode (in-house data), vehicle body deformation increased and occupant chest collision occurred that resulted in AIS4+ level injuries. This method of evaluation was confirmed to offer possibilities for evaluation of small overlap and oblique frontal crashes that conforms to actual accidents, and for evaluation that may further reduce fatalities and injuries. The effectiveness with respect to these injuries of measures taken through restraint devices and the vehicle body by means of numerical simulation using a human body model was verified. The results confirmed that combining vehicle body and restraint device measures had the effect of reducing injuries.