Equitable distribution of services at the regional scale and achieving a balanced spatial structure in the region are among the most important goals of sustainable regional planning, especially in developing countries such as Iran. In this regard, one of the most important strategies and policies in regional spatial planning is the decentralization of development. The purpose of this study is to identify the most important urban areas of Fars province in order to determine the priorities of development and decentralization of the unipolar development of the province. To this end, the regional network modelling based on graph theory and the concepts of social network analysis are used to conduct the spatial analysis of Fars province. The data used in this study includes  the Fars province road network extracted from the OSM open source system, as well as the spatial information of the urban and rural areas of Fars province extracted from the website of the Statistics Center of Iran. Regarding research methods, Pandas Library and NetworkX Library were used in the Python programming platform to form the network graph and analyze the centrality indicators, while ARC GIS software was for final processing and visualization of data and information. According to thenetwork centrality indices and the location of the province cities, notwithstanding Shiraz as the main hub of development in the province, Zarghan, Sadra, Kavar, Lepui, Khane Zenian and Noorabad are the main development priorities. In addition, the Shiraz-Marvdasht and Shiraz-Khaneh Zenian roads were determined to be the most important transportation paths and communication corridors. The results and findings of this study showed that network-based modeling and the use of graph-based analytics (in particular, social network analysis techniques) can be useful and effective as new analytical methods in regional studies and planning.