Abstract: Considering that the accessibility of urban parks is affected by many environmental factors and undergoes dynamic change in time, a research method for the spatial-temporal accessibility of attraction-oriented urban parks (STAAOP) is proposed by establishing a park attraction evaluation model and introducing time-based real-time navigation data. Firstly, based on the mass-source geographic information data, a park attraction evaluation model is constructed from multiple perspectives using the entropy weight method; then, based on the attraction evaluation results and the time-based real-time navigation data, an accessibility measurement model is constructed using the two-step mobile search method; finally, the standard deviation of accessibility change results is calculated and the clustering regions and the abnormal regions of accessibility change are identified using the local spatial autocorrelation analysis method. The intensity of accessibility change is analyzed from the overall and typical area perspective. Taking Jinan city for example, the results of this study show that the proposed STAAOP method can better capture the dynamic changes of the accessibility of urban parks and achieve better results both overall and regionally. This method provides a new idea and reference for more accurate and real evaluation of the accessibility of urban parks.