The invention discloses a method and a system for automatically extracting duplicate data. The method comprises the following steps of enabling a system server to send an Http (hyper text transfer protocol) service request to a plurality of application servers through an Http protocol, enabling the application servers to perform logic processing after a request message is received, and returning an xml (extensive usable language) format system webpage source file; enabling the system server to obtain an xml format data packet containing data from the source file, intercept a usable xml format source code file according to the service requirement, and analyze and package the xml format file; enabling the system server to store the analyzed data from each application server according to the designed data structure; enabling the data server to utilize the database technology to de-duplicate the duplicate data. The method has the advantages that the duplicate data of multiple systems can be extracted to find the problem of data duplicating of system information, and furthermore, the data flow and service working duplicating problems are analyzed and found.