Based on detailed and accurate statistical data of the government and enterprises, this paper studied the spatial flow pattern and optimization of Sinopec's first distribution of refined oil of Jiangsu province. In this study, we used the methods of Geographical Network Analysis and Linear Programming to calculate the unit oil transportation cost, combined with the optimized scenario construction and the comparative analysis of distribution pattern of refined oil. Then, we systematically analyzed the spatial flow pattern and optimized configuration of first distribution of refined oil, and attempted to explore its geographical network structure in theory, in the perspectives of allocation benefits and spatial structure. The results show that: (1) The cost under the optimization scenario is reduced by about 30% compared with the actual situation in 2017, indicating that the primary distribution benefit of Sinopec Jiangsu refined oil is not fully optimized. (2) Theoretically, according to the principle of "nearby allocation" of refined oil, different refining enterprises distribute the refined oil with lower unit oil transportation cost within their respective optimization scenario boundaries. (3) In fact, due to the differences in refining capacity of refineries, imperfect product oil transportation channels and different distribution center oil depots, there are relatively optimized scenario boundaries or large or small actual oil transportation boundaries between different refineries, and there are some distribution center oil depots (mixed oil transportation regions) that simultaneously accept the product oil of different refineries. We hope to further reveal the basic laws of spatial flow of refined oil, and provide decision-making advice from the field of economic geography for the current national market-oriented reform of refined oil and the preparation of national pipeline company.