By analyzing the statistical data of railway accidents, the relationship of such meteorological factors as rainfall, wind, visibility and incident probability was established according to regression analysis and grey relational analysis. Five geological factors, namely, the type of rock and soil, loose material reserve, drainage area, average slope and vegetation coverage were selected as the assessment indexes. Railway geological conditions risk assessment system was established based on AHP and multidimensional scaling method. The risk matrix expression was used for reference to integrate the geological and meteorological factors. The accident probability-risk matrix was given and the railway disaster risk assessment model was established. With railway disaster risk assessment model as the core, railway disaster monitoring and warning system was established including network layer, system hardware layer, data resource layer, information processing layer, application layer, business layer and access layer. At the same time, the data were received from railway safety monitoring information systems, China Meteorological Data Sharing Service System and the related information system of railway bureau. With four levels of ultrahigh, high, medium and low to represent risk intensity, the results of disaster risk assessment were displayed on railway GIS map, which could provide direct and visual railway disaster warning information.