Landslide is the most common form of geological disasters. Its main incentive is rainfall, so landslide often occurs in areas with abundant rainfall or occurs in flood season. Due to natural environmental and geological characteristics of Nanjing City, landslide is a major geological disaster there. In order to forecast landslide hazards effectively and minimize the loss caused by the unexpected landslide hazards, this paper established the landslide hazards weather forecast information sharing platform. On the basis of the existing data of geological hazards-prone areas, we use historic landslide data of Nanjing, meteorological data (rainfall) and the model of geological hazard forecast to establish the method of landslide hazards weather forecast and determine the landslide hazards weather forecast technique in Nanjing City. Then a forecasting information sharing platform is developed. Based on the method and the technology process of landslide hazards forecast, this platform uses C# as the development language and applies the following techniques: the database technology, the ArcGIS Server technology, the.NET technology, the remote call technology, the partial refresh technology of AJAX and the map cache technology. It also merges the geographic information system functions and landslide hazards forecast business functions together. This platform can achieve a dynamic and scientific management of the basic data of landslide hazards and the real-time meteorological data. It also can combine the real-time rainfall data to achieve the prediction of landslide hazard in Nanjing, and then it publishes the landslide hazards information on line in real time. At the same time, it provides the functions of landslide hazards information retrieval and statistical analysis and so on. Ultimately we hope this platform can provide a decision support for the landslide hazards prevention and mitigation and references for other cities which need to build a landslide hazards prediction platform.