Side impact continues to be a significant source of vehicle occupant injuries and fatalities requiring an improved understanding of response sensitivity to occupant position and the potential for injury. The primary interaction between the occupant and vehicle in near-side impact is the intruding door and vehicle structure where contact with the intruding structures is observed to be the most frequent cause of injuries. Restraints and protective systems are tested for a specific driving position with a prescribed set of anthropometric test devices; however, studies have shown that the initial position and posture of the occupant within the vehicle may have an influence on occupant response. In this study, a finite-element human body model with a detailed representation of the thorax and an ES-2re anthropomorphic test device (ATD) model were integrated with a mid-sized sedan vehicle model to investigate occupant response from a moving deformable barrier side impact. All models were verified using established tests (pendulum, side sled) and physical test data available in the literature. A parametric study was undertaken to evaluate the effect of arm position and car door material properties on thoracic response, assessed using thoracic deflection and the viscous criterion. In the standard driving position, the human and ATD model responses were similar. The human body model demonstrated significant sensitivity, relative to the ATD, to different arm positions with a lesser sensitivity to the door interior properties. It was found that in side impact which has the characteristics of high energy and limited displacement, the arm aligned with the body can act as a significant source of load transmission to the thorax, increasing the potential for injury. This study demonstrates that the choice of occupant model, injury metric and occupant arm position plays an important role in evaluating side impact safety.