Aims Net primary production (NPP) is the input to terrestrial ecosystem carbon pool. Climate and land use change affect NPP significantly. Shrublands occupy more than 20% of the terrestrial area of China, and their NPP is comparable to those of the forests. Our objective was to estimate China shrubland NPP from 2001 to 2013, and to analyze its variation and response to climate change. Methods We used a Carnegie-Ames-Stanford Approach (CASA) model to estimate the NPP of six shrubland types in China from 2001 to 2013. Furthermore, we used Theil-Sen slope combined with Mann-kendall test to analyze its spatial variation and a linear regression of one-variable model to analyze its inter- and intra-annual variation. Finally, a multi-factor linear regression model was used to analyze its response to climate change. Important findings We found the annual mean NPP of China shrubland was 281.82 g·m~(-2)·a~(-1). The subtropical evergreen shrubland has the maximum NPP of 420.47 g·m~(-2)·a~(-1), while the high cold desert shrubland has the minimum NPP of 52.65 g·m~(-2)·a~(-1). The countrywide shrublands NPP increased at the rate of 1.23 g·m~(-2)·a~(-1), the relative change rate was 5.99%. The temperate deciduous shrubland NPP increased the fastest with a speed of 3.05 g·m~(-2)·a~(-1) and subalpine evergreen shrubland had a decreasing trend with a speed of -0.73 g·m~(-2)·a~(-1). Moreover, the other four shrublands NPP had a growing trend, only subalpine deciduous shrubland NPP did not change significantly. The response of NPP to climate change of different seasons varies to different shrubland types. In general, the NPP variation was mainly affected by precipitation, and the spring warming also contributed to it. The increase of countrywide shrubland NPP may promote its contribution to the regional ecosystem function.