To support future solar energy deployment in China, long-term changes in solar energy resources over China were investigated based on high-resolution dynamical downscaling simulations under three emission scenarios. First, an evaluation of model performance was conducted through comparison with station and ERAS data, which indicated that although consistent overestimation was found, the spatial distribution and intra-annual variation of solar radiation were captured well by ensemble simulation, showing added value over regions with complex topography as well. Then, the trends of the solar power output from photovoltaic (PV) systems during 2020-2099 were projected, characterized by an increase in east and central China, and a consistent decrease in the solar-energy-abundant regions (e.g., northeast China, the Tibetan Plateau, and northwest China) under the three scenarios. Except for under the low emission scenario, the regional averaged trends of PV were projected to decrease consistently over China in all four seasons, with significant magnitude apparent in both spring and winter. In addition, a more intense magnitude of decline was found under the higher emission scenario, with more subregions with a significant trend and higher intra-annual and inter-annual variability. However, it seems that the emission scenarios impact the spatial pattern only remarkably in the late century, however strengthened the decreasing magnitude in all future periods. In addition, results indicate decline in the future PV potential over most of subregions of China due to reduced radiation and increased temperature, but for the decrease over Tibetan Plateau the reduced wind speed also act as an important factor. Plain Language Summary The long-term goal of carbon neutrality in China involves new requirements for the development and utilization of solar energy. The present study was motivated by supporting future solar energy deployment at local scales. However, due to data limitations, relatively little effort has been made to project long-term changes in solar energy over different subregions of China. Accordingly, we investigated this based on a set of high-resolution dynamical downscaling simulations under different climate scenarios, and found that the higher emission scenarios reinforce the magnitude of the decline in solar energy and result in higher inter-annual variability. These findings suggest an increased need for constraining emissions for the benefit of solar energy. In addition, we found that radiation will likely be the most important factor related to future changes in solar resources over most of subregions under RCP2.6 scenario, however the effect of wind speed should not be ignored over Tibetan Plateau.