Under perceived risk such as an imminent pedestrian accident, basic human survival instinct can result in sudden involuntary reactions which eventually affect muscle activity, posture and location just before and during the pedestrian impact. Consequently, pedestrian kinematics, kinetics and injuries could differ from that of predicted using passive human surrogates such as cadaver, dummy or passive numerical models. This could have multiple consequences as the procedures currently used to design and evaluate the performance of safety systems are based on such passive responses. For example, it could lead to complete failure of a safety system due to an inadequate posture or to underestimation of an injury risk in lack of additional muscle forces. Furthermore, increase in frailty, functional disorders and decline in visual, cognitive, sensory and reactive abilities with aging could affect these reactions. This could be important to consider as the elder population is significant portion of pedestrian fatalities. Thus the objectives of this project are to investigate: (1) Is reaction of a subject placed in pedestrian accident situation able to modify the outcome of accident in terms of kinematics and injury risk? (2) Are age related changes in reaction behaviour (i.e. reaction time and muscle force capacity) important to consider in this context? This project will be separated in two major tasks i.e. experiment and computer simulations. Tests will be performed with human volunteers of two different age groups (elderly and young). Their postural and muscular reactions to sudden audio-visual perturbations will be captured. Tests will be done in a controlled lab setting using non-impacting simplified setups to avoid risk to volunteer. Test results will then be used as initial conditions for vehicle-pedestrian impact simulations. Multibody or finite element models will be used. Simulation results will be analysed to evaluate the effects of reactions on the crash outcomes