Net primary productivity (NPP) is an important parameter for evaluating ecosystem vegetation growth and a key indicator for assessing the quality of the ecological environment. This study examined the six provinces in the hub region of "the Belt and Road"(B&R) in northwestern China as research areas. NPP in the study area during 1982-2015 was estimated using a modified parameter CASA model with GIMMS NDVI data, vegetation type data, and meteorological data. The nonlinear characteristics of NPP changes were revealed with the M-K and EEMD methods, and the response of vegetation NPP changes to climate change at different time scales was explored. The results showed that:(1) The NPP values in the vegetation growing season generally showed an increasing trend with a linear growth rate of 0.718gCm^-2a^-2. In most areas, NPP values should continuously maintain the current trend of change in the future, especially in southern Inner Mongolia, the Qinghai-Tibet Plateau, and the edge of the Tarim Basin. (2) The dynamics of vegetation NPP were mainly divided into 3-and 6-year time scales, and a long-term increasing trend. In northern Shaanxi, northern Xinjiang, northern Ningxia, northern Gansu, northern Qinghai, and central and northern Inner Mongolia, NPP was dominated by 3-year scale changes, whereas it was dominated by long-term changes in northern Shanxi, southern Gansu, southern Ningxia, southern Xinjiang, and eastern Inner Mongolia. For different vegetation types, shrubs, grasslands, and farmland were dominated by 3-year time scale changes and long-term growth trends, whereas coniferous forests, broad-leaved forests, and mixed forests were dominated by changes in the 3-year time scale. (3) The relationship between NPP change and temperature and precipitation became increasingly significant with the increase of the time scales. At the 3-year time scale, the correlation between vegetation change and temperature and precipitation was not significant in most regions (P > 0.05). At the 6-year time scale, there was a positive correlation between NPP and precipitation, which expanded towards the south. The correlation of the alpine meadows in southern Qinghai shifted from negative to positive. Over the long-term time scale, the relationship between vegetation NPP and temperature and precipitation became very significant. In most study areas, areas with positive correlations were much larger than areas with a negative correlation. Our results highlight that multi-time scale analysis can better analyze the spatial and temporal characteristics of NPP and its response to climate change at different time scales, which can help reveal the nonlinear response mechanism of vegetation NPP to climate change under global warming and evaluate the eco-environment induced by climate change. Our data provide scientific support for the sustainable development and ecological environment protection of the six provinces in the northwest.