To explore the influence of various factors on the spatial differentiation of urban heat island intensity, and to reveal the impact mechanism of the factors, spatial expression, spatial overlay, and geographical detector methods were used in this study. The impact of five influencing factors on the spatial differentiation of urban heat island intensity in the study area was examined, and a multivariate relationship model was constructed. The results show that the intensity of urban heat island in the Guangdong-Hong Kong-Macao Greater Bay area is high in the central part and low in the surroundings, which has formed a semicircular urban heat island belt on both sides of the estuary of the Pearl River. According to the results of the geographical detector analysis, the five selected factors have a high explanatory power on the spatial differentiation of the urban heat island intensity at the 1 km×1 km grid scale, in the order of population density (0.668) > proportion of construction land area (0.577) > length of roads (0.573) > proportion of vegetation cover (0.538) > proportion of surface water area (0.428). The constructed multivariate relationship model can accurately reflect the distribution of land surface temperature in urban heat island area, and the error between the modeling result and the observed average land surface temperature is 0.34℃.