In response to the current problem of only single standard to evaluate the meter in China and a lower accuracy in its life prediction under complex environment stress, a deep research on the life prediction of smart meters under complex environment stress is conducted. The failure mechanism of smart meters is studied through the analysis of 6 stresses existing in environment which affect smart meter’s life. In addition, a smart meter’s life prediction solution based on T-S（Takagi-Sugeno） fuzzy model with particle swarm optimization is proposed to address the difficulty of predicting lifespan due to the small number of fault samples and long lifespan of intelligent energy meters. Then stress and lifespan data of smart meters under four typical climatic conditions in China are collected for experimental verification of lifespan prediction, and experiments about the life prediction are made through these data. The experimental results show that the proposed approach is an effective way to predict the life of smart meters under different environments and various environmental parameters, with higher accuracy than the original T-S fuzzy model algorithm.