Taking 6 provinces including Jiangxi, Hubei, Shangdong, Jiangsu, Guizhou and Tibet for example,the paper used information entropy theory and principal component analysis to open out the variation characters of water structure as well as the impact mechanism of driving factors in different areas in China from 13 factors which involves four aspects such as economy, society, hydro meteorology, agriculture and water resources.It also analyzed the effect of the most stringent water management system on water use structure.The results showed that (1) In the past 16 years, the total amount of water used in Shandong Province in east China continually declined, and its information entropy of water use structure was at a high level with an obvious rising trend; the total amount of use water in Tibet in west China increased markedly, and its information entropy was at a low level. There is no big change of the total amount of water use in other areas, and 91 percent of water is production water supply on average. (2)The production water has gradually decreased since the most stringent water management system was published, and the information entropy of water structure in each province has a rising trend. (3) The main factors driving the variation of water structure in these areas are industrial output value, population, and local water resources amount and total water consumption.