A data merging method for distributed computing uses a configuration file to guide the insertion of computational results obtained by distributed nodes into a database table, and to merge the computational results inserted in the database table. The configuration file is established according to the task splitting conditions of distributed computing. The method uses a data merging server to import and host the configuration file and create a database table corresponding to the configuration file. Each distributed node inserting its distributed computational result into the database table after completing the respective distributed computing task, while the data merging server merges the data in the database table. An interface standard applicable to various uses and various distributed computing tasks may be used such that the user only needs to establish a job-specific configuration file based on the interface standard.