Under the background of rapid urbanization, an in-depth understanding of the spatiotemporal variation characteristics of urban park systems is crucial for rational park allocation and sustainable ecological environment development. Using the central urban area of Beijing as a case study, we employed the Arc Geographic Information System（ArcGIS） and Fragstats platforms to analyze the spatial pattern and service level quantitatively. We also explored the overall features and partition disparities of the spatiotemporal changes in the urban park system from 2000 to 2020 and revealed their policy-driven factors. The results showed that the urban park system in the central urban area of Beijing has expanded continuously, with an average annual growth rate of 18.45% in the number of parks from 2000 to 2020. This growth exhibits four stage characteristics: rapid growth (2000–2005), more rapid growth (2005–2010), steady growth (2010–2015), and accelerated growth (2015–2020). The evolving park system tends towards complexity and diversification. The most significant direction of the dynamic change of the system varies across different stages, with the alterations between the fourth and sixth ring road being particularly prominent. Concurrently, the increase in the number of parks facilitates the continuous enhancement of the system’s service capacity while also aggravating the repetitive allocation of parks to some extent. Although the emphasis has shifted from quantity growth to rational layout development, there are still numerous park service blind areas and severe duplicate configurations in the city. Service levels also vary significantly among diverse zones, indicating unbalanced development across the system. Notably, the repetitive allocation situation of parks in some zones has exacerbated again in recent years, so it is urgent to pay attention to the scientific layout of parks. The priority of policy implementation is the main cause of these spatiotemporal differences in the system, which promotes the differentiated increase in the park number and subsequently leads to the unbalanced development of the system. Finally, in this paper, we proposed corresponding optimization strategies as a reference for the accurate allocation and fine management of the park system.