Aiming at the problems existing in the operation and management of long-distance water diversion projects, such as heterogeneous data of monitoring and testing, various types of data, nonstandard management and low efficiency, a data management system based on classification tree was constructed by analyzing the types and acquisition methods of multi-source data. Combined with database technology, the database table structure of the management system was constructed, in order to optimize the classification management of multi-source data. Based on this, a multi-source data management system for long-distance water diversion projects in operation period was established based on B/S (browser/server) architecture. This system was designed in strict accordance with the data management system with clear hierarchy and strict structure, which facilitates the unified and efficient management of data by system users, and provides a new solution for multi-source data management during the operation period of water diversion projects. At present, the system has been put into use in a water diversion project with satisfactory operation effect.