The Loess Plateau in China has experienced a significant change of landscapes and vegetation since the implementation of Grain to Green Programme. Based on a comparison of the changing trends of the fractional vegetation cover (FVC) and the net primary production (NPP) at three temporal scales (i.e. 2000-2005 a, 2000-2010 a, and 2000-2015 a), this paper formulated a composite index to assess the effectiveness of vegetation restoration. Effectiveness together with its relationship with socio-economic factors was explored via the structural equation modeling (SEM) for three time periods. The spatial non-stationary factors for vegetation restoration, including climatic and key socioeconomic indicators based on SEM were then explored using the geographically weighted regression (GWR). The effectiveness of vegetation restoration demonstrated the remarkable improvement of vegetation restoration, covering 88.20% of the Loess Plateau. The high value area was concentrated in the northern Shaanxi and Shanxi province. The relationships between effectiveness and socio-economic factors indicated that the strongly negative impact of population pressure was alleviated (the path coefficients was -0.95 and -0.86, respectively) since the rural employment decreased. Moreover, improvements in the rural economy may provide the possibility of improving the restoration effectiveness, which should particularly focus on improvements in agricultural practices. The climate and socio-economic factors that affected the effectiveness of vegetation restoration had significantly spatial differences on the Loess Plateau. The effectiveness of vegetation restoration on eastern the Loess Plateau, covering 29.30% of the area was dominated by the mean annual precipitation, with the positive influence. The mean annual temperature was the leading factor of the effectiveness of vegetation restoration on the northern and western sandstorms of the Loess Plateau, accounting for 20.93% of its total area. Whereas, the central and western parts of the Loess Plateau, covering 47.02% of its total area were more affected by socio-economic factors. Our study revealed the continuous improvement of vegetation restoration and its key climatic and socioeconomic factors, which supports the promoting strategy for the optimization of vegetation restoration policies.