Thermal infrared (TIR) remote sensing techniques have been used over urban areas and applied in urban climate and heat environmental studies for more than 30 years, mainly for analyzing land surface temperature (LST) distribution, difference and their relationship with city development and surface energy fluxes. The rapidly expanding urban areas of the world form an environmental challenge for the 21st century that requires both new analytic approaches and new sources of data and information. In developing countries, however, remote sensing may provide large-scale, fast and repeated observations of urban growth and environmental conditions that are usually not available from other sources or with high cost. Because of the rapid urbanization, large numbers of the natural surfaces in the city have been replaced by artificial surfaces. This has resulted in undesirable thermal impacts on city meteorological and ecological environment-Urban Heat Island (UHI), especially bringing lots of negative effect to the health of human beings and living environments. Some studies have showed that the green land and water areas (mainly in city parks), can just diminish UHI effect on temperature and increase air moisture through the photosynthesis, transpiration and evapotranspiration of vegetation around urban parks. In this paper, 17 Guangzhou City parks were selected as the case areas and the temperature information was inversed from Landsat TM remote sensing data. After analyzing the temperature distribution patterns surrounding the 17 parks using field investigation, buffer area analysis and cubic polynomial fitting method, we find that the temperature differences between the surrounding sites and boundaries of park rise with the distances further away from the park boundaries. However, the temperature accretion rate is step-down at the same time. The shape of the temperature increasing curve is similar to a cubic polynomial through the origin of coordinates (R2 > 0.84). The maximum cooling range model of city parks in Guangzhou was built. The maximum cooling ranges are from 14-432m; and the maximum cooling values of temperature are at about 1.9-4.3K. What is more, the average temperature decreasing range around parks has a positive relationship with the green area of parks, and the fitted curve is similar to that of a logarithmic. For those parks that the water-covered-rates are larger than 50%, their temperature cooling effects are always more remarkable. For those parks that the length-width-rates are more than or equal to 2, the temperature cooling effects are also more obvious, even if their park areas are smaller than others. From the viewpoint of the cooling effect of parks to their surrounding environment in Guangzhou, the optimal values of planned green land should be between about 4200 m2 and 540000 m2. Therefore, the park area and distance from boundary of a park played the significant role in changing the effect of UHI. This study is the first quantitative research on the temperature distributions of main parks′ surroundings in Guangzhou City, which may be of the directive significance to the ecology planning (especially city parks) of large cities like Guangzhou City. It also provides a reference and theory basis for other cities to design and plan the city parks in the world.