Abstract: By 2023, a comprehensive Chinese forage dataset had been constructed based on forage phenological stages, plant height, and yield observations archived at the National Meteorological Information Center, covering the period from 1981 to 2023. This dataset had undergone rigorous data cleaning, quality control, and verification procedures. It encompassed decadal records of forage phenology (annual), plant height, and yield from 302 agrometeorological observation stations across 14 provinces (autonomous regions and municipalities) in China. The dataset provided critical observational support for climate change research, containing records for 110 forage species in phenology, 99 in plant height, and 82 in yield. Among these, 53 forage species with continuous observations spanning more than five years exhibited average data availability rates of 77.7%, 96.7%, and 98.5% for phenology, plant height, and yield, respectively. Following stringent quality assurance protocols, the overall data accuracy for these 53 species exceeded 96.0% across all three variables. Specifically, phenological data achieved the highest accuracy at 99.1%, surpassing that of plant height (96.1%) and yield (96.9%). Notably, 23 species—including Poa crymophila, Artemisia frigida, and Agropyron repens—demonstrated perfect (100%) accuracy in their phenological records.