To provide a scientific basis for the policies for the development of wind energy and towards the goal of carbon neutrality in China and local governments, changes in wind energy potential over China were investigated based on the bias-corrected ensemble mean (CENS) of high-resolution dynamical downscaling projections using the RCP2.6, RCP4.5 and RCP8.5 emission scenarios. Firstly, evaluation indicated that CENS could better reproduce the long-term mean and interannual variability of surface wind speed over China compared to the original simulation outputs, providing a reliable basis for future projections. Projection of the averaged wind power density of China showed a decreasing trend during 2020–2099 in all four seasons under all three scenarios, with the significant (p < 0.05) magnitude apparent in both autumn and winter, characterized by an increase over South China and a decrease over northwestern and southwestern regions. In addition, the ‘very abundant’ and ‘abundant’ categories of available wind energy located in northern China and the low-speed wind energy in South China were projected to decrease by the middle and late of the 21st century. Although the projected decrease in annual wind power density ranged from −1.14% to −0.36% per decade among the three scenarios, we believe that, considering the strong inter-annual variability and uncertainty involved, these changes would not substantially affect China's future wind energy plans. Also of note was that a low emission scenario could to some extent mitigate the reduction in wind energy potential in the future. Furthermore, we suggest that implementation of newly developed wind power technology should consider the impact of changes in wind energy in different sub-regions (e.g., the low wind speed region over South China), which is pivotal to China's strategic planning in this sector.