Abstract: 【 Purpose/significance】The government is the master of massive data, and measuring the efficiency of government data open⁃ ness and exploring effective paths for data openness are conducive to improving the speed of realizing the value of government data and enabling data. 【Method/process】Taking the efficiency perspective as the starting point, a super-efficiency SBM model was con⁃ structed, and the data open efficiency values of 16 representative cities were measured, and then the efficiency values were used as the result variables to explore the effective path of government data openness using fsQCA.【 Result/conclusion】The results show that gov⁃ ernment data openness is affected by many complex factors, and from the perspective of condition combination, there are six paths for high-level government data openness, which are manifested in four modes: government-led, economically supported, internal and ex⁃ ternally linked, and all-factor-driven.【 Innovation/limitation】This paper takes the efficiency of government data openness in the three years from 2019 to 2021 as the starting point, integrates information technology factors, and provides a path reference for government data openness. However, the efficiency of government data openness is affected by many aspects, and the conditional variables se⁃ lected in this article may also have certain limitations.