Shanghai was used to systematically evaluate the cooling effectiveness of urban parks from the perspective of thermal comfort to address the limitation of existing studies that rely heavily on the land surface temperature （LST） and neglect the thermal comfort in environment.The modified temperature-humidity index （MTHI） was introduced to characterize the thermal comfort.The vector data of parks were integrated with mobile signaling data to evaluate the cooling capacity in parks from the perspective of thermal comfort.The cooling benefits of parks and urban-rural disparities based on MTHI and LST were compared.The results showed that the cooling range of parks increased from the core urban area to the outskirts， but the benefits evaluated with MTHI was overall lower than those with LST.The size of park， vegetation coverage， and water-body proportion was positively correlated with the intensity of cooling.The vegetation coverage was positively correlated with LST-based cooling distance， but negatively correlated with MTHI-based cooling distance.The equivalent cooling service for population in the core urban area far exceeded that in the suburbs due to its high density of population.However， the served population estimated with MTHI was 54.5% lower than that with LST， indicating that traditional LST-based evaluation may significantly over-estimate the benefits of parks.It is indicated that the thermal-comfort indexes should be introduced to comprehensively evaluate the cooling benefits of parks and adopt differentiated strategies including increasing the supply of small parks in high-density urban areas and optimizing the design of parks in suburban areas to accurately improve the thermal comfort of the living environment.