The dataset with 2.5 arcminutes (~4 km) was spatially interpolated based on 699 weather stations in China using thin-plate spline and random forest methods with six covariates (e.g., location, terrain, and reanalysis data), including daily 2m maximum temperature (maxtmp), maximum temperature (mintmp), mean temperature (meantmp), total precipitation (pre), skin temperature (gst), 10m wind speed (win), relative humidity (rhu), surface pressure (prs), and sunshine duration (ssd). The dataset covers the mainland area of China and ranges from 1 January 2000 to 31 December 2020. The dataset was well evaluated by independent weather stations and reliable for the study related to climate change across China. Zhang Jielin, Liu Bo, Ren Siqing, Han Wenqi, Ding Yongxia, Peng Shouzhang. A 4 km daily gridded meteorological dataset for China from 2000 to 2020. Scientific Data, 2024, 11, 1230. https://doi.org/10.1038/s41597-024-04029-x