Abstract: The urbanization process and population growth have led to the fragmentation of biological habitats and the loss of biodiversity. It is of great significance for habitat restoration and biodiversity protection to use effective models and indicators to evaluate landscape connectivity and build green space network. Taking Fengtai District of Beijing as an example, this research firstly applies the connectivity index based on the graph theory to explore the optimal distance threshold of green space construction suitable for biological migration, and selects the source patches according to the evaluation results. Secondly, it uses the least cost path model to construct the resistance surface, and the Linkage Mapper tool to determine the potential connection path of species migration. Finally, it identifies the “pinch point” area which is vital for species migration by quantifying the relative importance of the source patches and corridors with the circuit density, and compares the results with the remote sensing satellite map and empirical observation of bird abundance. The results show that the ecological base of the green space in the west region is good, which provides the main habitat environment for the species. The green patches in the central and eastern regions are seriously fragmented. Therefore, it is necessary to increase the urban green space in the pinch areas as stepping stones for species migration. In this research, the circuit model focusing on species diffusion is introduced to make up for the limitations of other green space network construction methods, and clarify the current situation of habitat quality and the future development of green space network in Fengtai District of Beijing, which can provide scientific reference for the optimization of regional green space patterns and the formulation of biodiversity conservation planning strategies.