In order to better understand the thermal comfort of urban parks and provide empirical reference for urban green space optimization design, 5 days’ field monitoring was conducted in People’s Park in Urumqi, an oasis city in an arid region of China. Combined with GIS spatial interpolation, correlation analysis, and regression analysis, the spatial and temporal distribution of thermal comfort (HI and WBGT) of urban parks was discussed. The results showed the following. (1) The thermal comfort in the morning was generally higher than that in the afternoon, and the thermal comfort near the water body and lush vegetation in the park was higher, while the thermal comfort on the road was lower, especially on Hotan Street and Binhenan Road, which were far away from the park. Therefore, it is recommended that nearby residents exercise outdoors in the morning as much as possible and in the park, and in the afternoon, keep to the park and its vicinity and try to sit quietly or walk slowly, avoiding the less comfortable areas, such as Hotan Road and Binhenan Road. (2) Due to dense vegetation and lack of infrastructure construction, the thermal comfort area does not have the conditions for crowd gathering. Therefore, it is recommended that the park improve the infrastructure of relevant areas. (3) Through the analysis of the significant influence of explanatory variables on the explained variables, it shows that the ventilation effect in the park is insufficient. Therefore, it is recommended to appropriately increase the number of trees, water bodies, and wind channels to promote ventilation in the park so as to improve the thermal comfort of the park. These findings provide a theoretical basis and technical reference for optimizing the thermal comfort of urban green space and establishing a healthier and more comfortable living environment for urban residents.