The Loess Plateau (LP) of China stands out as one of the world's most prominent greening hotspots. While previous studies have largely attributed this greening to ecological restoration initiatives such as 'Grain for Green Program (GFGP)', it remains unclear whether the observed increase in vegetation greenness is primarily due to land use changes related to GFGP or to vegetation restoration within unchanged land use types. To address the knowledge gap, this study utilizes high-resolution (30 m) vegetation greenness and land use data to quantitatively assess the contributions of different greenness baselines and land use types to the overall greening of the LP from 2000 to 2020. Our findings revealed a significant upward trend in vegetation greenness across the LP during the study period. Notably, stable grasslands with lower-greenness baselines emerged as the largest contributor to regional greening, accounting for 47.6% of the observed increase, followed by stable croplands at 36.8%. Ecosystems with greenness-baseline values between 0.2 and 0.5 played a pivotal role in driving the greening process, with the most substantial increases occurring in areas characterized by lower-greenness baselines. In contrast, regions with higher-greenness baselines exhibited relatively modest improvements. Importantly, this study demonstrates that the greening of the LP was predominantly driven by the enhancement in grasslands and croplands with lower-greenness baselines, rather than by land use conversions such as cropland-to-forest or cropland-to-grassland transitions. By highlighting the disproportionate contribution of lower-greenness-baseline areas, this research provides scientific evidence and novel perspectives to inform the protection and restoration of degraded ecosystems at regional and national scales. The increase in vegetation greenness on the Loess Plateau (LP) of China was previously largely attributed to ecological restoration initiatives such as 'Grain for Green Program (GFGP)'. Using utilizes high-resolution (30 m) vegetation greenness and land use data over 2000-2020, this study finds that the greening of the LP was predominantly driven by the enhancement in grasslands and croplands with lower-greenness-baselines, rather than by land use conversions such as cropland-to-forest or cropland-to-grassland transitions. (sic)(sic): (sic)(sic)(sic)(sic)(sic)(sic)(LP)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic) "(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(GFGP)" (sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(30(sic))(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)2000-2020(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)((sic)(sic)(sic)47.6%), (sic)(sic)(sic)(sic)(sic)((sic)(sic)(sic)36.8%).(sic)(sic)(sic)(sic)(sic)(sic)0.2-0.5(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).