Abstract: Under the background of severe decline in global biodiversity, the enhancing and protection of biodiversity has become a focus of current urban green space research. The habitat network plays a key role in the biodiversity protection, and the optimization of urban habitat network can more effectively support the protection and restoration of urban biodiversity. Taking the central area of Beijing as the target, this research innovatively couples MSPA and InVEST models to select habitat source areas, giving comprehensive considerations to the spatial morphological and functional attributes of patches, based on which the habitat corridors and habitat nodes are identified. The research indicates that the MSPA-InVEST model has advantages and feasibility in the optimization of green space habitat network at the central city scale: the connectivity of large-scale habitat patches is deficient, the habitat is fragmented and shows a trend of polarization in the habitat quality; 205 habitat sources (32 core habitat sources), 463.2 km habitat corridor (88.5 km key habitat corridor) and 527 habitat nodes (49 significant habitat nodes) are identified in the research area. The scale, quality and connectivity of the habitat network should be further enhanced in the future. Under the background of ecological construction in the new era, the research results can accurately guide the construction of green space in the central area of Beijing in the future, and provide important support for the protection and restoration of biodiversity.