Energy efficiency in residential buildings can be improved considerably by optimizing the allocation of thermal resistance (R) and capacitance (C) of external walls. However, it is difficult to identify which RC allocation yields the minimum heat loss or heat gain for a given climate region, due to a lack of understanding of how regional climate affects optimal RC allocation. In this study, we determine the effect of the climate characteristics on the development of optimal RC allocation by combining the RC network model and the particle swarm optimization algorithm, taking five representative climates in different Chinese cities as examples. The energy efficiency potential and the ratio of heating to cooling load are introduced to comparatively analyse the energy savings after optimization for the five cities. The results indicate that the energy efficiency potential of residential buildings in Harbin is the largest, ranging from 6.2% to almost 20.6% for different total building capacitance. In Harbin, Beijing and Kunming, the energy efficiency potential is larger than those in Guangzhou and Shanghai, which is attributed to the larger heating to cooling load ratio of the former than that of the latter. An approximately linear relationship between the energy efficiency potential and the ratio of heating to cooling load is also built to fast predict the maximum energy savings for any given climate regions. This study could serve as useful references for the optimal design of thermal resistance and capacitance in different climate regions to save energy of residential buildings.