The invention discloses a whole-network-domain electronic transformer high-frequency measurement data discrimination and data processing method. It carries out the high-frequency measurement data discrimination in a mode of combining a Leida criterion and a clustering analysis method, and carries out the screening and removing of abnormal data; based on the high-frequency measurement data, it adopts a Spark offline parallel computing mode to discriminate the high-frequency measurement data; and stores the discriminated data by adopting an HDFS-based distributed operation data storage mode. According to the method, by adopting an efficient data discrimination algorithm, useless data in the high-frequency measurement data is effectively screened out, and a good data basis is provided for further data analysis; a Spark parallel computing technology is adopted, so that the speed is greatly improved, and the discrimination efficiency of the high-frequency measurement data is effectively improved; and the screened high-frequency measurement data is stored by adopting HDFS-based distributed operation data, so that the data storage and access efficiency is improved.