Abstract: With the continuous expansion of oil and gas stations in construction scale in China, the requirements for safe production, operation and management are becoming more and more stringent. However, the visualization application and comprehensive display of current pipelines and station entities are very scarce, and it is impossible to view the regional distribution and operation data of such stations in a three-dimensional scene. In addition, facility and equipment models highly consistent with the actual situations have not yet been established. In response to those problems, a systematic solution of intelligent oil and gas stations based on digital twin was proposed. Specifically, according to the digital twin theory, an equal-scale high-precision station model was established on the basic platform, i.e. ArcGIS system, by the 3Ds Max software to build a digital twin-based station that could truly map the physical station. Based on that, an intelligent cyber-physical system of oil and gas stations was created with the help of advanced sensing, big data analysis and human-computer interaction technology. As a demonstration, the system design and development was applied to a compressor station in Shaanxi, proving that the system could realize the three-dimensionality display of all elements in the station, the real-time monitoring of station status, and the overall decision-making on station management. Hence, it is helpful to comprehensively improve the level of macro supervision, major supervision and precise supervision of stations for safe production.