Assessing the elevation-dependent alpine grassland phenology is of great significance to reveal the relationship between vegetation and climate change. Based on the two Normalized Difference Vegetation Index (NDVI) from SPOT and MODIS and the Enhanced Vegetation Index (EVI) from MODIS during the period of 2000-2013, DEM, eco-geographical and grassland distribution map, the patterns of elevation-dependent alpine grassland phenology over the Tibetan Plateau (TP) were analyzed. The results showed that: (1) With an increasing elevation gradient, the distribution of spring and autumn phenology in alpine grassland exhibited a regular change, while the magnitude of the trend of spring phenology in alpine grassland was quite different from that of autumn phenology on the TP. (2) There was a dividing line of approximately 3200 m above sea level on the trends of spring phenology in alpine grassland with an increasing elevation gradient. Below this line, the magnitude of spring phenology advancement increased significantly, whereas above the line, it decreased significantly. However, the magnitude of autumn phenology delay in alpine grassland decreased and then the magnitude of the advancement increased significantly across the altitudinal gradient. (3) Opposite trends of alpine grassland phenology change were observed compared with southern Qinghai Plateau alpine-meadow-steppe region (ICl) and southern Tibet montane shrub-steppe region(IICl). Changes in the trends of spring phenology with increasing elevation were quite different between the above two regions, but those of autumn phenology were synchronized. These differences suggest that the vertical variability of the alpine grassland phenology derived from remote sensing could not directly mirror the elevation-dependent warming on the TP.