Dry and wet division is one of the most basic contents in climate classifications. In order to explore regional potential features, 20 global climate models (GCMs) were statistically downscaled to reproduce temperature and precipitation at a resolution of 0.25° × 0.25° across China, which illustrated agreeable performance in comparison with observation. Taking temperatures as inputs, the Hargreaves model was implemented to estimate potential evapotranspiration (PET). This model was typically recalibrated to keep accuracy in further usage, which resulted in normalized root mean squared error being less than 5%. The indicator defined by the ratio of annual precipitation to annual PET, namely dry/wet index (I_DW), was projected in potential dynamic of dry/wet division. I_DW was consistently expected in an increasing trend under RCP4.5 relative to the baseline period of 1986–2005. Regards RCP8.5, I_DW was diagnosed in a decreasing trend in northern Xinjiang and most central-southern regions, but an increasing trend in most northern regions, implying dry tendency in current wet condition in southern parts while wet tendency in dry condition in northern parts. The possible contribution of precipitation and PET could unveil regional differential change in I_DW. It was highlighted that the area exposed to extreme dry, and dry was likely to decrease, but the exposure to wet and extreme wet tended to increase in the future. These can provide a better knowledge of potential change of climate and water resource, supporting adaptive strategies in response to climate change.