The invention provides a method and system for evaluating the running state of an intelligent electric energy meter. The method comprises the following steps: firstly, constructing a plurality of Bayesian networks based on historical data; calculating the similarity between the Bayesian networks, selecting a plurality of Bayesian network individuals with high precision and strong diversity, and finally integrating the selected Bayesian networks by adopting a convex function evidence theory method to obtain the probability value of each operation state of the intelligent electric energy meter.According to the method and the system, the state evaluation of the intelligent electric energy meter is accurately realized, and great convenience is provided for the safe operation, state monitoring, data reading, electricity price issuing and other work of a power grid, so that the huge cost of periodically recovering and detecting the electric energy meter is effectively saved, and a powerfultechnical guarantee is provided for replacing the intelligent electric energy meter.