Future changes of climate extremes in the 21st century over Xiongan New Area and Jing-Jin-Ji district were investigated based on high resolution (6.25 km) combined statistical and dynamical downscaling datasets, which were produced using the observation of CLDAS, five sets of regional climate change simulations by RegCM4.4, and statistical downscaling with quantile mapping. Of that, the RegCM4.4 simulations were conducted over East Asia under RCP4.5 scenario driven by five different CMIP5 global climate models of CSIRO-Mk3-6-0, EC-EARTH, HadGEM2-ES, MPI-ESM-MR and NorESM1-M. Validations of the present climate show that the multi-model ensemble mean can well reproduce the spatial distribution of most climate extremes, and better performance can be found in the temperature-related climate extremes. However, some biases can also be observed, especially for the consecutive drought days (CDD). In the context of global warming, increased extreme warm events, decreased extreme cold events and consecutive drought days and increased extreme heavy precipitation events are projected in Xiongan New Area and the whole Jing-Jin-Ji district. In specific, increased TXx (Maximum value of daily maximum temperature) and TNn (Minimum value of daily minimum temperature) can be found, with the value exceeding 2.4℃ and 3.2℃, respectively. More pronounced increase of SU (Number of summer days) over the mountainous areas compared with the plain is observed, while greater increase of TR (Number of tropical nights) is found over the plain. The increase of SU and TR are in the range of 20 ? 40 d and5 ? 40 d, respectively. Both the FD (Number of frost days) and ID (Number of icing days) will decrease, with the decline above 10 d and 5 d, respectively. Precipitation-related climate extremes including CDD, R_1mm (Annual count of days when daily precipitation≥1mm) and R_10mm (Annual count of days when daily precipitation≥10 mm) are mainly on decrease with small values of ?10% ? 0 while increase of R_X5day (Maximum consecutive 5-day precipitation), SDII (Simple precipitation intensity index) and R_20mm (Annual count of days when daily precipitation≥20 mm) are found in most areas with the values in the range of 0 ? 25%. Regional mean changes show that the linear trends are more significantly in temperature-related climate extremes compared with those in precipitation-related climate extremes. Comparing the two regions, greater uncertainty of the simulations in Xiongan New Area can be found, which indicates the deficiency of the model in local scale areas.