Climatic and anthropogenic disturbances play pivotal roles in shaping the dynamics of grassland. Quantifying their impacts on grassland variation is essential to ensure sustainable grassland management. In this study, we employed the Thornthwaite Memorial and Carnegie–Ames–Stanford‐approach (CASA) models to investigate the spatiotemporal effects of these two variables on grassland variation in northern China from 2000 to 2016, using the net primary productivity (NPP) as a measure. Our findings reveal that approximately 25.92% of the grassland in northern China experienced degradation, while restored grasslands occupied 45% of the total grassland area. The average grassland actual NPP (ANPP) and human‐induced NPP decreased at rates of −0.60 and −5.62 gC m−2 a−1, respectively. Conversely, potential NPP exhibited an upward trend with an average increase of 2.27 gC m−2 a−1. Furthermore, grassland ANPP showed a projected increase in most parts of northern China. Climate change emerged as the primary driver for grassland restoration in Xinjiang, Qinghai, and Inner Mongolia, leading to an increase of 21582.79 Gg C in grassland NPP. In contrast, human activities were the dominant catalysts for grassland degradation, resulting in a reduction of 51932.3 Gg C in grassland NPP. Human‐induced grassland degradation was ubiquitous in northwest and northeast China. With the exception of slope grassland, climate change primarily influenced the restoration of most grassland types, while human activities were the primary cause of degradation. Our analysis indicated a strong correlation between temperature and grassland degradation, while precipitation played a pivotal role in grassland restoration in northern China. Human interference demonstrated both positive and negative impacts on grassland changes. In conclusion, the increase in precipitation and the implementation of ecological restoration plans have effectively promoted the restoration of grasslands in northern China.