The China Annual Tree Cover Dataset (CATCD) is the first long-term, high-spatial-resolution annual tree cover dataset for China. It is generated by integrating time-series Landsat imagery with ensemble learning techniques based on random forests. The reliability and accuracy of CATCD have been rigorously validated against multi-source reference data (Correlation: 0.70–0.96; RMSE: 5.6%–25.2%). The current release extends the 30-meter resolution annual tree cover dataset for China up to the year 2025. Data for previous years can be accessed from earlier versions of this repository. Each year's data is stored as a GeoTIFF file containing two bands: 'TreeCover' and 'Uncertainty'. The 'TreeCover' values range from 0 to 100, representing the percentage of tree canopy cover within each pixel. The 'Uncertainty' band provides an estimation of tree cover uncertainty—expressed as the standard deviation of predictions from five models—derived through 5-fold cross-validation. Beyond mapping annual tree cover, users can leverage CATCD to extract forest dynamic ranges by applying specific tree cover thresholds. In addition to Zenodo, CATCD can be accessed, downloaded, and analyzed on the Google Earth Engine (GEE) platform via the following link: https://code.earthengine.google.com/366132348ace002478a2256a71022f74. Please note that while the Zenodo repository hosts the annual tree cover products, the corresponding uncertainty estimates are exclusively available on GEE. CATCD is scheduled to be updated annually at the end of each year across both platforms. Data Citation: Yaotong Cai, Xiaocong Xu, Sheng Nie, Cheng Wang, Peng Zhu, Yujiu Xiong, and Xiaoping Liu (2024). Unveiling Spatiotemporal Tree Cover Patterns in China: The First 30m Annual Tree Cover Mapping from 1985 to 2023. ISPRS Journal of Photogrammetry and Remote Sensing, 216: 240-258. DOI: 10.1016/j.isprsjprs.2024.08.001. Contact: For data-related inquiries, please contact Dr. Yaotong Cai (caiyt33@mail2.sysu.edu.cn).