Based on WSR system methodology, this paper uses fuzzy set qualitative comparative analysis method to analyze the conditional configuration of data governance cases of 31 provincial governments in China for studying the theoretical support and implementation path of Chinese government data governance. The study found that government data governance is comprehensively influenced by physical dimension, reasonable dimension and human dimension, among which physical dimension and reasonable dimension play a central role. The transformation process of government data governance can be divided into physics-oriented, reason-oriented, human-reason combined and three-dimension mixed. The core conditions of the four configurations have made different emphases and each auxiliary condition plays a different role in the course. Because of the different path types, China’s government data governance model has new characteristics of innovation, complexity, comparability and coordination. In order to build a digital government subject that meets the demand of data governance, the practical application of government data governance can be accelerated from three aspects including infrastructure guarantee, system guidance and people first.