As an essential green infrastructure in the city, the green park space plays a huge role in protecting biodiversity, reducing the urban heat island effect, and providing leisure and entertainment places for the residents. Whether the spatial allocation of green park space can reasonably meet the needs of residents and realize the resource guarantee will be an urgent problem to be solved in the current research of urban green park space planning. Therefore, with the green park space of 11 streets in Quanshan District of Xuzhou City as the research object, the ArcGIS network analysis function is used to evaluate the supply fairness of the two-step mobile search method, Gini coefficient, and supply and demand relationship, and the optimal layout scheme of the green park space is proposed. The study found that: (1) the spatial layout of green park space in the study area is unbalanced, 75% of the community supply is less than 14.6 m2/person, the service scope has a blind area; (2) Population density and park green space layout show space mismatch, the Gini coefficient of one street exceeds the warning line, and four streets are close to the warning line;(3) The supply and demand difference in the low supply and high demand area, the six streets in the study area belong to the low supply and high demand area, and the supply and demand relationship needs to be improved; (4) In terms of layout optimization strategy, the three-step optimization method is adopted to optimize the allocation of 243.79 hm2 park green space within 10 years, so that the 92 communities with insufficient supply can reach the standard, and the community-level Gini coefficient is reduced by 47%. The research results provide data support for optimizing the park green space system in Xuzhou City and improving the fairness of the park green space service supply.