Industry is a critical part of economy, and also a major source for carbon emission. This paper uses calibrated gravity model and social network method to analyze China ‘s 2005 to 2019 industrial carbon emission, and applies QAP to explore its factors. The overall network features suggest a rising spatial connection among provinces, who need to collaborate thoroughly toward energy-saving-emission-reducing. Eastern provinces/cities such as Jiangsu, Zhejiang, Shanghai and Tianjin are positioning in the centers of social networks with a more complicated connection, less difficulties in connecting other provinces and controlling more resources, while the central and western provinces are on the contrast. The eastern coastal provinces are at the centers, with their inner connection in the core higher than in the margin, but growing rate lower, suggesting an increasing inner connection inside the marginal areas. QAP regression results show that the five variables, industrialization, technology, energy intensity, industrial structure and energy industry, can promote spatial connection of industrial carbon emission from their variances. This paper presents suggestions on boosting regional cooperation, realizing regional collaboration, accelerating green transformation in terms of social network features and SAP regression.