AbstractVegetation phenology is a sensitive indicator of climate change and has significant effects on the exchange of carbon, water, and energy between the terrestrial biosphere and the atmosphere. The Tibetan Plateau, the Earth's “third pole,” is a unique region for studying the long‐term trends in vegetation phenology in response to climate change because of the sensitivity of its alpine ecosystems to climate and its low‐level human disturbance. There has been a debate whether the trends in spring phenology over the Tibetan Plateau have been continuously advancing over the last two to three decades. In this study, we examine the trends in the start of growing season (SOS) for alpine meadow and steppe using the Global Inventory Modeling and Mapping Studies (GIMMS)3g normalized difference vegetation index (NDVI) data set (1982–2014), the GIMMS NDVI data set (1982–2006), the Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI data set (2001–2014), the Satellite Pour l'Observation de la Terre Vegetation (SPOT‐VEG) NDVI data set (1999–2013), and the Sea‐viewing Wide Field‐of‐View Sensor (SeaWiFS) NDVI data set (1998–2007). Both logistic and polynomial fitting methods are used to retrieve the SOS dates from the NDVI data sets. Our results show that the trends in spring phenology over the Tibetan Plateau depend on both the NDVI data set used and the method for retrieving the SOS date. There are large discrepancies in the SOS trends among the different NDVI data sets and between the two different retrieval methods. There is no consistent evidence that spring phenology (“green‐up” dates) has been advancing or delaying over the Tibetan Plateau during the last two to three decades. Ground‐based budburst data also indicate no consistent trends in spring phenology. The responses of SOS to environmental factors (air temperature, precipitation, soil temperature, and snow depth) also vary among NDVI data sets and phenology retrieval methods. The increases in winter and spring temperature had offsetting effects on spring phenology.