Abstract. Monitoring global radiation resources relies on sunshine duration (SD) as a significant indication, but there is a scarcity of research that have examined high-resolution SD data. This study established a daily 5-km SD dataset in China from 2016 to 2023 using Himawari's Advanced Himawari Imager (AHI) Level 3 shortwave radiation fitted with the Ångström-Prescott model based on time series. We used ground-measured SD at 2380 Chinese Meteorological Administration (CMA) stations to verify the accuracy of SD dataset. The results of the testing set indicated that the average correlation coefficient (R) between the SD from estimation and the ground-measurement is 0.88. We investigated the effects of wind speed, vapor pressure (VAP), precipitation and aerosol optical depth (AOD) on the estimated performance of SD, and the results showed that temperature had the greatest effect on SD estimation. We also found that both too low AOD and too high wind speed also affected the SD estimation. This high-resolution SD data can provide important support for accurate radiation resource assessment in China. The SD dataset is freely accessible at https://doi.org/10.57760/sciencedb.10276 (Zhang et al., 2024).