Due to the rapid development of modern construction activities, the landslide disaster induced by artificial disturbance is increasing gradually. Therefore, landslide monitoring is becoming more and more important to ensure the safety of human and stability of engineering structures. An improved system is introduced to monitor and to pre-warn the whole process of large deformation landslide disaster. This system based on two-way communication platform includes Beidou satellite and general packet radio service (GPRS). The zigbee technology is applied to build a wireless multi-hop self-organizing sensor network. The new anchor cable with constant resistance and large deformation functions, which can absorb energy, is used to replace the common anchor cable as the transmission device. Thus two-meter deformation of slope rock mass is allowed without tensile failure of energy absorption anchor cable. In addition, four level warning criteria are established according to four color grades (i.e. red, orange, yellow and blue). 3D automatic searching and processing software system is also developed. Vector maps, such as topography remote sensing map, geologic profile map and distribution map of monitoring points, are embedded in the monitoring area. Therefore, this new system has benefits in: searching area automatically, processing data intelligently, displaying monitoring curve automatically, and querying monitoring information quickly. The system has been used in Nanfen open pit iron mine. According to the warning criteria, the warning information of the landslide on October 5, 2011 was sent out 5 days in advance. Combined with the monitoring data of rainfall and the cumulative amount of mining, it is proved that sliding force is the only effective parameter for landslide pre-warning, which results from the coupling analysis of the evolution of the various parameters before and after the landslide. The successful application of this system provides theoretical and practical basis for monitoring and pre-warning other similar slopes.