AIDS is an infectious fatal disease caused by HIV, which is class B in infectious disease in China. Since the first case of AIDS reported in 1985, AIDS has been rampant in China. Exploring the spatial pattern of AIDS and its spatiotemporal evolution characteristics will help improve AIDS prevention and control. In this study, we adopted GIS spatial statistical methods to analyze the provincial incidence data of AIDS in China from 1997 to 2016. First, we used spatial autocorrelation technology to detect the spatial pattern of the AIDS epidemic. Then, we explored the spatiotemporal evolution process by using the centroid transferring curve model. Results show that: (1) The epidemic of AIDS in China has strong spatial dependence at the provincial scale. From 1997 to 2016, the global spatial correlation of AIDS increased, and is likely to further increase. The development and diffusion process of the AIDS accorded with the first law of geography. (2) The AIDS epidemic in China showed a general pattern of "high in the south, low in the north, and random in the middle." The regions where local spatial autocorrelation occurred could be divided into two areas: the northern low-low clusters represented by Inner Mongolia and the southern high-high clusters represented by Guangxi. The cold spots area of AIDS in the north experienced fluctuations and increased slightly, and extended to the northeast and central China. The hot spots area in the south had a growing trend. (3) The overall prevalence of the AIDS epidemic has been expanding from 1997 to 2016 gradually, with obvious regional differences. In the process of diffusion, thespatial pattern of AIDS became increasingly unbalanced. Our findings suggest that, to achieve regional synergy and precise control of AIDS in the future, in addition to the traditional prevention methods, we should focus more on the spatiotemporal patterns of AIDS diffusion. Also, it is necessary to strengthen the control of hot spots in the epidemic and the direction of high-risk transmission. The present study demonstrates the importance and reliability of the spatial statistical analysis methods in improving medical and health services, and could be used as a scientific reference for the work of AIDS prevention and control in China. In future studies, we should scale down the research unit when more detailed data is available.

Keywords： China ; AIDS ; spatial autocorrelation ; center of gravity model ; spatial pattern ; spatiotemporal evolution