The impressive visual sensitivity to human movements is illustrated by the ability to recognize actions even when the whole body movements are reduced to the movements of a few point-lights. Across 3 experiments, we investigated the role of the conditions of visualization and ownership in perceptual weight judgment tasks. The observers were displayed minimal information on movements consisting of either the working point displacement only or additional configural information on the kinematics of the trunk, shoulder, arm, forearm, and hand joined by a stick diagram. Furthermore, the participants were displayed either their own movements or those of another participant while lifting different objects’ weight. The results revealed that although participants’ overall performance was not dependent on the conditions either of visualisation (working point vs. stick diagram) or ownership (self vs. other), the kinematic cues used to perform the task (i.e., mean acceleration, maximal height, maximal velocity, duration of the phases of movement: reach, grasp, lift, and place) differed as a function of these conditions. In addition, the results revealed that the kinematic parameters that are relevant for action observation do not entirely match the ones that are relevant for action execution. We discuss the implications of these results for our understanding of action observation in terms of the opposing motor simulation and visual analysis hypotheses.