This paper investigates the spatial distribution of inbound and domestic tourist flows to cities in China and their growth rates using exploratory spatial data analysis. This method is a set of GIS spatial statistical techniques that are useful in describing and visualizing the spatial distribution, detecting patterns of hot-spots, and suggesting spatial regimes. The global Moran’s I statistics for inbound and domestic tourist flows reveal strong positive and significant spatial autocorrelation. Furthermore, the Moran significance maps indicate four significant inbound tourism hot-spot areas in 1999 and 2006 (the Beijing-Tianjin cluster, the Yangtze River Delta cluster, the Fujian coast cluster and the Pearl River Delta cluster), and five significant domestic tourism hot-spot areas in 2002 and 2006 (with the addition of the Chengdu cluster). Based on the results, we show that tourism flows are polarized into clusters and remain very stable over time. As has been seen in other countries, hot-spots tend to be in insular and coastal areas, and associated with cities that are higher in China’s urban hierarchy. Implications are drawn, such as priority of resource allocations for hot-spot area and the utilization of spillover effects from hot-spots.