【Objective】As urbanization progresses, the urban heat island effect intensifies gradually. Studying the relationship between surface temperature and urban green spaces is important for mitigating the urban heat island effect, as it is a common factor in characterizing the urban thermal environment. However, there are few quantitative studies on the spatial and temporal variations of surface temperature in green spaces alone, and there is a lack of quantitative analyses of the four main types of surface temperature variation (constant, conversion, increase and loss) in different types of urban green spaces at a macroscopic scale. 【Method】 We used four phases of Landsat5 TM and Landsat8 OLI remote sensing image data from Beijing every six years between 1999, 2005, 2011 and 2017. We inverted the surface temperature of the area within the Sixth Ring Road of Beijing and simultaneously performed a remote sensing interpretation at the same time. We then separately extracted the surface temperature of the green space to analyze its spatial and temporal evolution characteristics. 【Result】 Calculating the normalized surface temperature values within the Sixth Ring in 1999, 2005, 2011 and 2017 revealed some differences in surface temperatures in different substrata. For example, the surface temperature of arable land is generally located in the urban fringe area and is less affected by urban construction and anthropogenic heat. Therefore, its surface temperature is lower than that of the green spaces in central urban areas. We observed some cooling characteristics in the green spaces, whereas the remaining substrates showed warming characteristics. The different components of the substratum were one of the main reasons for differences in urban surface temperatures. The spatial distribution of green spaces within the Sixth Ring Road changed to some extent from 1999 to 2017. The overall transfer of green space was 227.09 km². Except for grassland, the other three types of green space tended to be transferred, and the highest proportion was transferred to impermeable surfaces. This continuous urban expansion and construction has led to changes in the composition of the urban substrates, resulting in increased surface temperature and intensification of the urban heat island phenomenon. The spatial and temporal evolution of green space leads to changes in surface temperature. An increase in green space leads to a decrease in surface temperature, whereas the exchanged, lost, and unchanged parts of green space show different degrees of warming. The loss of green space contributed more to the increase in the normalized surface temperature than the exchange of green space. For green space exchange, except for the conversion of grassland to other types of green space, which had a cooling effect, the remaining conversion types showed some warming effects. 【Conclusion】 To mitigate the urban heat island effect and create a more suitable living environment, increasing the area of green space, reasonably optimizing its layout and structure, protecting areas of permanent basic agricultural land, and converting grassland into other types of green space under the guidance of Beijing’s policy of “Leaving white space and increasing green space” is necessary. In the future, monitoring the urban thermal environment at different micro levels can be further combined to form a multi-scale and multi-dimensional surface temperature research mechanism, increase the level of surface temperature research, and provide a more diversified basis for subsequent research alongside macro-scale research on surface temperatures in urban construction areas.