Localization is one of the most important issues in wireless sensor networks (WSNs). In the most widely proposed range-free algorithms, nodes estimate location by employing the geometric constraints imposed by the location of the mobile anchor. However, none of them addresses how the mobile anchor moves to optimize the improvement of location inaccuracies and minimize the anchor's movement. This paper assumes that previous range-free algorithms have been executed for a period of time and the deployed sensors are of different location inaccuracies. According to the size of the estimative region of each static sensor, an anchor-guiding mechanism is proposed to determine the beacon locations and construct an efficient path for the mobile anchor. Experimental study reveals that the proposed anchor-guiding mechanism effectively guides the mobile anchor to move along an efficient path, thereby saving the time required for improving or balancing the location inaccuracies of all sensor nodes.