The banking industry plays an essential role in China's economic growth, and the assessment of bank efficiency has attracted increasing interest in the past two decades. Given that the data envelopment analysis cross-efficiency method can provide a more reliable reference set, a large number of studies targeting performance evaluation and its applications adopt this method. Few studies, however, address the issue of the non-uniqueness of weights in cross-efficiency evaluation. Whereas, this paper proposes a cross-efficiency interval considering all weighting schemes along with a VIKOR aggregate model to gauge the universal productive efficiency of 16 major Chinese commercial banks of China and rank them for the period of 2007–14. Specifically, we use the cross-efficiency interval to generate the global performance of each bank based on internal and external assessment while using VIKOR to compare the solution with the attitude index which ranks all of the banks. The results indicate that the cross-efficiency interval can provide more information than the traditional DEA model, that the banking system efficiency of China has been improved during the study period, especially for SOBs; that JSBs are more efficient compared with SOBs; and the efficiency gap narrowed during the study period.