The invention relates to a 
system and a method for assessment of walking and miming 
gait in human. The method is preferably based on the fusion of a portable device featuring inertial sensors and several new dedicated 
signal processing algorithms: the detection of specific temporal events and parameters, 5 optimized fusion and de-drifted integration of inertial signals,  automatic and online virtual alignment of sensors module, 3D foot 
kinematics estimation, a kinematic model for automatic online 
heel and 
toe position 
estimation, and finally the extraction of relevant and clinically meaning-full outcome parameters. Advantageously including at least one 
wireless inertial module attached to foot, the 
system provides common spatio-temporal parameters (
gait cycle time, 
stride length, and stride velocity), with the 10 
advantage of being able to work in unconstrained condition such as during turning or running. It furthermore may provide original parameters for each 
gait cycle, both temporal (load, foot-flat and push duration) and spatial (foot clearance and 
turning angle), and their inter-cycles variability. The 
system and method according to the invention allows the assessment of various aspects of gait which have shown recently to be of premium importance in research and clinical field, including foot clearance, 15 turns, gait 
initiation and termination, running, or gait variability. The system may be light weight, easy to wear and use, and suitable for any application requiring objective and quantitative evaluation of gait without heavy laboratory settings.