We studied the changes of vertical contact forces, lower limb kinematics, and electromyographic activity (EMG) at different speeds and gravitational loads. To this end healthy subjects were asked to walk on a motorized treadmill while the percentage of body weight unloaded (body weight support, BWS) was modified in steps by means of a well-characterized unloading system. BWS was set at 0, 35, 50, 75, 95, or 100% of body weight. Walking speed was 0.7, 1.1, 2, 3, or 5 km/h. We found that changing BWS between 0 and 95% resulted in drastic changes of kinetic parameters but in limited changes of the kinematic coordination. In particular, the peak vertical contact forces decreased proportionally to BWS; at 95%-BWS they were 20-fold smaller than at 0% and were applied at the forefoot only. Also, there were considerable changes of the amplitude of EMG activity of all tested lower limb muscles and a complex re-organization of the pattern of activity of thigh muscles. By contrast, the corresponding variation of the parameters that describe shape and variability of the foot path was very limited, always <30% of the corresponding values at 0 BWS. Moreover, the planar co-variation of the elevation angles was obeyed at all speed and BWS values. Minimum variance of limb trajectory occurred at 3 km/h. At 100% BWS, subjects stepped in the air, their feet oscillating back and forth just above but never contacting the treadmill. In this case, step-to-step variability of foot path was much greater than at all other BWS levels but was restored to lower values when minimal surrogate contact forces were provided during the "stance" phase. The results did not depend on the specific instruction given to the subject. Therefore we conclude that minimal contact forces are sufficient for accurate foot trajectory control.