This paper presents the value of "smart" restraint systems in mitigating the injury risk to occupant's in a greater range of impact conditions than those presently considered in current regulatory and consumer impact tests. The work was carried out under the European 5th framework project PRISM (Proposed Reduction of car crash Injuries through improved SMart restraint development technologies). A generic MADYMO compartment model of a typical European midi- MPV was developed with a conventional restraint system for the modelled driver. To identify variables that need to be considered in the performance of a "smart" restraint system and subsequently assess potential adaptations that could be made to a restraint system, two simulation studies were carried out with the developed model. The first of these studies investigated the influence that the following variables have on driver injury risk: occupant size (using both a Hybrid-III and human body models), the reclined position of the seat, the bracing response of the driver and thoracic fracture. Based on the models' predictions it is implied that the kinematics and predicted injury risk of various sizes of human model are very different from those of a 50th percentile Hybrid-III dummy and that the reclined position of the seat and bracing response of the driver increases injury risk. It was not clear if fractures in the thoracic region would contribute to an increase in injury risk to other body regions. In the second simulation study investigations were performed to assess alterations that could be made to the modelled restraint system to adapt its performance to better protect different occupant sizes. It was concluded that if it were possible to adapt restraint characteristics to the specific occupant size, injury risk could be lowered.