The forest canopy height (FCH) is a crucial indicator of ecosystem processes, significantly influencing biomass distribution and biodiversity. Accurately and efficiently obtaining large-scale FCH information is vital for monitoring forest health and carbon storage. However, current research and products predominantly offer single-time measurements of FCH, with limited attention to its temporal dynamics. This study integrates ICESat-2 Lidar data with Landsat time series (1985-2020) to construct 30-meter resolution mean and maximum FCH datasets for the Loess Plateau, assessing spatiotemporal dynamics and growth potential. The results indicate that our FCH product shows good agreement with ground validation (Mean FCH R2 = 0.744, RMSE = 3.993; Maximum FCH R2 = 0.705, RMSE = 4.535). A significant increase in FCH was revealed from 1985 to 2020, particularly after 2015, with notable growth in the southern region. Annual FCH growth rates ranged from 0.1 to 0.2 m/year, particularly in semiarid areas. The potential for canopy height growth is highest for the southern forests of the Loess Plateau, and the magnitude gradually decreases with increasing latitude. This study offers valuable insights for optimizing ecological engineering efforts in the region and contributes to improving the long-term effectiveness of restoration strategies.