The invention provides a power network element loss abnormity determination method, for solving the problems existing in conventional power network element loss abnormity determination. The determination method comprises the following steps: selecting an actual measuring day to carry out a power network element line loss rate test; calculating a statistical line loss rate and a theoretical line loss rate of a power network element selected on the actual measuring day, and obtaining a difference between the statistical line loss rate and the theoretical line loss rate; and setting a power network element loss abnormity threshold K, when the absolute value of the difference between the statistical line loss rate and the theoretical line loss rate is smaller than K, determining that the loss of the selected power network element is normal, and otherwise, determining that the loss of the selected power network element is abnormal. According to the invention, a conventional determination mode according to line loss rate fluctuation conditions counted at the end of a month and empirical values is abandoned, the disadvantage of lagged management is avoided, the problem is effectively solved that power network element loss abnormities are conventionally determined based on the empirical values, which is not suitable for the requirement for line loss refinement management, and the method provided by the invention has quite high economic benefits and a quite high practical value.