The indoor thermal conditions of residential buildings during the plum rain season in East China (Nanjing for example) were obtained and the accuracy of three different building energy consumption models (whole building heat, air and moisture model HAM, conduction transfer function model CTF, effective moisture penetration depth model EMPD) were analyzed. The numerical model was implemented into Matlab-Simulink and was validated using measured data from a field survey. Then the validated model was used for simulation analysis using typical weather data of plum rain season.The results from the survey showing the indoor temperature was higher than 28℃ and the indoor relative humidity was higher than 70% in most time of the plum rain season of 2013. Simulation results showed that there was not much difference when using the three models for thermal simulations, however, using CTF model for moisture simulation would result in large inaccuracy. During the plum rain season in Southeast China, when the room ventilation rate is less than 2 ACH, the moisture buffering effect of building envelope is obvious and the choice of proper hygroscopic materials can save up to 30% of building energy consumption.