In recent years, the Chinese government has made great efforts to jointly control and prevent air pollution, especially fine particulate matter (PM2.5). However, these efforts are challenged by technical constraints due to the significant temporal and spatial heterogeneity of PM2.5 across China. In this study, the Affinity Zone Identification Approach (AZIA), which combines rotated principal component analysis (RPCA) with revised clustering analysis, was developed and employed to regionalize PM2.5 pollution in China based on data from 1496 air quality monitoring sites recorded from 2013 to 2017. Two clustering methods, cluster analysis with statistical test (CAST) and K-center-point (K-medoids) clustering, were compared and revised to eliminate unspecified sites. Site zonation was finally extended to the municipality scale for the convenience of the controlling measures. The results revealed that 17 affinity zones with 5 different labels from clean to heavily polluted areas could be identified in China. The heavily polluted areas were mainly located in central and eastern China as well as Xinjiang Province, with regional average annual PM2.5 concentrations higher than 66 mu g/m(3). The new approach provided more comprehensive and detailed affinity zones than obtained in a previous study (Wang et al., 2015b). The North China Plain and Northeastern China were both further divided into northern and southern parts based on different pollution levels. In addition, five affinity zones were first recognized in western China. The findings provide not only a theoretical basis to further display the temporal and spatial variations in PM2.5 but also an effective solution for the cooperative control of air pollution in China. (C) 2020 Elsevier Ltd. All rights reserved.