Building intelligent oil fields is the only way for the high-quality development of the petroleum industry in the future. At present, oilfield enterprises mainly adopt a top-down data traceability based on business in the process of data governance, as well as the compilation of enterprise data resource catalogues containing common data definitions. This approach has achieved certain results in improving the accuracy of data for individual systems, but due to the independent construction of different systems, the governance of a single system has not taken into account the continuity and integrity of the actual overall business. Therefore, there are still serious problems with data quality defects in the overall business. In response to the above difficulties, this article innovatively proposes a “one source and four dimensions” data governance model, which prioritizes the authenticity of data through upstream and downstream system correlation and virtual reality business interaction. The practical application in oilfield production capacity construction business scenarios has shown that this method can not only effectively improve the overall data quality of production and construction business, but also synchronously promote the compliance of business processes at various nodes in production and construction project management, achieving dual improvement of data and management level in oilfield production and construction. It has good promotion and reference significance in enterprise data construction and data governance engineering.