A spatial correlation network of grey water footprints in China’s provinces from 2000 to 2014 is constructed by gravity model. The structural characteristics and effects of spatial correlation network of provincial water pollution are analyzed from the perspective of social network analysis. The results show that the inter-provincial grey water footprint has obvious network characteristics. The spatial connection became more and more close, the hierarchical network structure gradually loosened, and the network tended to be stable from 2000 to 2014. Shanghai, Tianjin, Jiangsu, Beijing, Zhejiang and other provinces are at the center of the network and play an important “hub bridge” function in inter-provincial spatial association. The network can be divided into four functional blocks: main benefit, net benefit, broker and main overflow. The inter-plate spillover effect is obvious with the characteristics of reciprocity, tripartite transmission and circular transmission. The geographical spatial distribution of the plates shows the characteristics of large agglomeration and small dispersion. There are two spatial correlation modes between the plates: geographical proximity and cross-regional boundary. The network structure has a significant impact on the intensity of grey water footprint. The improvement of network density and node centrality, the reduction of network grade and network efficiency can effectively promote the decline of grey water footprint intensity, and reduce the inter-provincial differences.