Generally, the failure distribution characteristics of different failure modes are different, and the stress types and degradation processes that induce their failures also differ. This paper proposes a method for predicting the life of electric energy meter based on Multi-stress influence Weibull distribution model. The stress types mainly include environmental stress and electrical stress. Using historical fault data of electric energy meter to establish Weibull distribution model of failure rate of each failure mode, and then for the main types of influence stress, establish a model between the influence stress and failure rate of each failure mode, and calculate the influence coefficient. The short-term life of the overall energy meter is obtained by adjusting the failure rate at the predicted stage of each failure mode. By verifying the field actual energy meter fault data, showing that the method can effectively predict the life of the energy meter, and provide technical support for energy meter status replacement, and fault early warning.