This report details a project to review and develop the JARI pedestrian model as chosen by the IHRA to develop their pedestrian head impact test procedure. In the work several modifications were made to the model, including the removal of duplicated contacts defined between the model’s arms and legs, modifications to allow axial stretching in the spine, and the implementation of shoulder joints. To test the biofidelity of the simulated shoulder joints predictions from the original and modified versions of the JARI model, in addition to those from a pedestrian model developed by TNO, were compared against the results from shoulder impact studies completed on PMHS’s. It was found that all the models demonstrated very poor shoulder biofidelity. Furthermore, for the same simulated vehiclepedestrian impacts, differences in the predicted head impact velocities and head impact angles from all the pedestrian models were as high as 3.9 m.s-1 and 17.1° respectively.