Abstract: Landslide hazard is influenced by many temporal and spatial factors.Traditional spatial analytical techniques cannot easily discover new and unexpected patterns,trends,and relationships that can be hidden deep within very large diverse geographic datasets.Focusing on Three Gorges Reservoir Area,environmental and triggering factors for landslide occurrences were extracted from multi-source data.Then,quantitative landslide susceptibility indices were calculated using the trained three-layered BP neural network,and the landslide susceptibility maps were generated.Finally,success rate curve was used to verify the results of landslide susceptibility mapping,and the results showed the best accuracy of 89.75%.The validation showed sufficient agreement between the prediction results and existing landslide.Therefore,the proposed model is an efficient method for landslide intelligent prediction,and can provide a significant reference for landslide hazard prediction and assessment.