Abstract: Net primary productivity (NPP) and its response to global change are two of the focuses in global change research. Based on MODIS NDVI data, land use classification data, and meteorological data, the spatio-temporal change in NPP of Inner Mongolia grassland was simulated using an improved light use efficiency model (CASA model). The relationships between NPP and climate factors were analysed based on partial correlation coefficients of the two-group elements. The average annual NPP from 2001 to 2010 was 281.3 g C/(m²·a) in the study area, with an obvious increasing trend from southwest to northeast. The mean NPP in the unit area of meadow steppe was 431.8 g C/(m²·a), in typical steppe it was 288.7 g C/(m²·a) and in desert steppe was 123.5 g C/(m²·a). There was a rising trend in the grassland NPP during the past ten years. Grassland NPP in Mu Us sandy land, Hunshandak sandy land, Horqing sandy, HulunBuir league and the southern slope of the Greater Hinggan Mountain exhibited the most obvious increasing trend, while in the Yin mountain region and the centre of Xilin gol league it showed the most obvious decreasing trend. As a whole, precipitation was the dominant driving force of NPP of Inner Mongolia grassland, while the NPP of meadow steppe correlated with both precipitation and temperature, but with more dependence on temperature. Compared with temperature, precipitation influenced NPP of typical steppe and more seriously, desert steppe.