AbstractSpring vegetation phenology (green‐up onset date, GUD) exhibits notable sensitivity to climate change, serving as a critical indicator of ecosystem dynamics. However, long‐term changes and drivers of GUD remain unclear. Here we showed that satellite‐derived GUD averaged over China forests and grasslands advanced by −1.3 ± 0.4 (mean ± SD) days decade−1 during 1982–2022, but with contrasting trends between forests (−5.0 ± 0.6 days decade−1) and grasslands (2.8 ± 0.6 days decade−1), despite similarly increasing temperature and precipitation. Such contrasting trends were caused by different responses to higher preseason mean temperature and more total precipitation. Moreover, sensitivities of GUD to temperature and precipitation showed different patterns with respect to spatial gradient of background climate conditions between forests and grasslands. Our study elucidates different mechanisms behind forests and grasslands responding to climate change, which could help optimize land‐management strategies and anticipate vegetation distribution under climate change.