In the ZigBee-based positioning system, the positioning signals used to estimate the distance are mainly by detecting their received signal strength indicator (RSSI) or link quality indicator (LQI). Due to the RSSI easily performed and no additional hardware needed, there are many various RSSI measuring methods proposed by the previous studies in order to create the environment models of the relation between RSSI and distance. As using the environment model to perform positioning procedure, the various measure methods result in different error degrees of the positioning accuracy. Therefore, a low performance complexity and high positioning accuracy measure method will be proposed, named cutting ceiling-floor method, to create the environment model. The experiment results show the good positioning performance by the proposed method. Moreover, in the most researches, they select three highest RSSI level nodes as references from the all fixed nodes to perform the positioning procedure. It usually obtains the worse positioning results. Therefore, a new method to select reference nodes appropriately is proposed according to the slope of curve in the environmental model. When the new selection method is operated to estimate the target positions, the experiment results show that it effectively reduces the erroneous range and improves the positioning accuracy in comparison with selecting the highest RSSI level nodes as the references. Finally, adopting both the cutting ceiling-floor and the reference node selection methods for the various positioning system, the experiment results show that they always have obviously improvements for positioning accuracy in the different environments. As a result, the two proposed methods are useful for the ZigBee positioning systems to promote the performance. Keywords： ZigBee, RSSI, LQI, Positioning.