Chronic bilateral internal globus pallidus (GPi) stimulation allows control of levodopa induced dyskinesias (LID) and motor symptoms in severe Parkinson's disease (PD). The effect on gait has not been clearly established. Different results have been reported, mostly consisting of clinical data. The aim of this study was to evaluate, by means of a video motion analysis system (optoelectronic VICON system), the influence of bilateral GPi stimulation on gait in PD. Five patients underwent bilateral GPi stimulation. The preoperative and postoperative (3 months after surgery) clinical gait disturbances (items 29 and 30 of the motor UPDRS), as well as spatial and temporal gait measurements (namely cadence, velocity, stride and step times, single and double limb support times, stride and step lengths) were analysed in off condition (the patient had received no treatment for 12 hours or merely the lowest dose of levodopa allowing him to walk for the gait analysis) and in the on drug condition (after administration of 200 mg of levodopa). The gait analysis was performed with the VICON system. In off condition, there was a statistically significant improvement after surgery for UPDRS III and gait (clinically assessed). In on drug condition, there was a significant improvement for LID whereas UPDRS III and clinical assessment of gait were unchanged. The VICON system also showed that surgery improved gait especially in off condition, but also in on drug condition. Our method allowed exact quantification of the influence of surgery on gait characteristics. As compared with levodopa treatment, the effect of stimulation seems to be different. Indeed, the results suggest only limited effects of pallidal stimulation on the control of stride length and rather point to compensatory additional mechanisms.