The invention relates to an analysis method for electricity stealing behavioral mode of user electricity utilization of a user. The analysis method comprises the steps of: (1) acquiring electricity utilization data of the user and acquiring anomalous event data of an ammeter; (2) extracting historical electricity utilization data and historical anomalous event data of the ammeter for processing; (3) cleaning the historical electricity utilization data and the historical anomalous event data of the ammeter; (4) establishing an electricity stealing behavioral analysis model; (5) correcting electricity stealing model parameters; (6) evaluating an electricity stealing risk. The analysis method is carried out under a framework of a big data platform, and mass data are calculated in parallel, so that the problem of low calculation efficiency is solved, the requirements on daily analysis, daily processing and daily monitoring of an electricity utilization department are satisfied, a clerk per se can select an electricity stealing user in high risk for processing, and therefore the work efficiency is improved.