Abstract: Rapid urbanization has significantly altered the original urban spatial configuration, exerting profound implications for land surface temperature (LST). However, the dynamic mechanisms by which urban spatial expansion influences LST remain poorly understood. Focusing on the Changsha-Zhuzhou-Xiangtan urban agglomeration of China, this study employs multi-source remote sensing data and landscape pattern analysis to examine changes in urban spatial configuration from 1990 to 2020, including land cover, landscape pattern, and three-dimensional morphology. Furthermore, it investigates the effects of these changes in urban spatial configuration on LST. The findings are as follows: 1) between 1990 and 2020, impervious surfaces within the region continued to expand at the expense of cropland and green spaces. Landscape patches shifted from a dispersed to an aggregated pattern, and high-rise buildings proliferated in central urban areas. 2) Regions with high LST were predominantly observed in the urban center, corresponding to the accumulation of impervious surfaces, and the urban heat island effect continuously strengthened over the study period, resulting in the formation of multiple heat island centers. 3) The contribution rates of land cover and landscape pattern indicators progressively declined, whereas those of three-dimensional morphology indicators steadily increased. Among these indicators, the contribution rate of building height rose from 3.45% in 2000 to 12.19% in 2020, making it an important factor influencing LST. As urban areas approach spatial saturation, optimizing three-dimensional urban morphology emerges as the most effective strategy for mitigating urban heat island effects. This study clarifies the evolving mechanisms linking urban spatial configuration to LST, providing a scientific basis for urban spatial planning and heat island mitigation in rapidly urbanizing agglomerations.