To improve the understanding on response mechanism of the vulnerable ecological system on the Tibetan Plateau to the climatic change, relationship between alpine grassland vegetation index and ground surface temperature from 1982 to 2006 is investigated over the Tibetan Plateau. The data resources are ground surface temperature (GST) data from meteorological stations and relative cell NDVI values quoted from the GIMMS NDVI database. Employing regression analysis method, correlations between alpine grassland′ NDVI and GST by various time-scale and the annual changes both of them are focused on. The results show that both of alpine grassland NDVI and GST showed positive trend from 1982 to 2006 overall the plateau. The trend of annual mean NDVI, NDVI in growing season, maximum NDVI and annual mean GST, GST in growing season were 0.007, 0.011, 0.007 per decade and 0.60, 0.43 degrees Celsius per decade, respectively. The area of maximum NDVI associated with GST accounted for 70.49%. However, spacial heterogeneity in topography, climatic condition and hydrological condition led the complexity in the impact of GST to the growth of grass. The correlation between maximum NDVI and annual mean GST was found to be most significant; in reviving and withering period, NDVI and GST were significantly positively correlated. The correlation of the maximum NDVI to the GST clearly depended on bio-coverage: they briefly negative correlated for poor vegetation cover or degrading regions; but positive for the good vegetation cover.