Countries around the world attach great importance to the performance of government data governance. However, existing research has not established a systematic analysis framework, and the interpretation of government data governance performance is limited. Therefore, from the perspective of institutional theory, public value theory and resource-based theory, this paper constructs an analytical framework for value goals, institutional environment and government capacity, and uses data service utilization variables, policy environment support, public demand pressure, government financial capacity and Leadership is highly valued as a condition variable. The fuzzy set qualitative comparison method is used to analyze the configuration of 31 provincial-level government data governance cases in China. The study finds that the performance of government data governance is affected by multiple factors, including three types of leadershipdemand, system-goal, and comprehensive driving. The class combination path promotes high performance of government data, while the non-high performance 3-class generation mechanism is completely different.