The combined impact of global climate change and accelerated urbanization has intensified the urban heat island (UHI) effect, heightening population heat exposure risk. This study integrated MODIS LST (Land Surface Temperature), LUCC (Land Use Cover Change), GUB (Global Urban Boundaries), DEM (Digital Elevation Model), and population data to identify the spatial expansion of UHI in China from 2005 to 2020. A population-weighted urban heat island intensity (UHII) index (Ep) was proposed based on gridded UHII and population to compare the population heat exposure risk for the newly developed and pre-existing UHI under three levels of UHII. In 2020, UHI areas doubled since 2005, with infilling type exhibiting the highest UHII, while edge-expansion type was largest, and leapfrogging type had the lowest UHII. In the summer of 2020, the Ep indices in China were 1.75 °C and 1.79 °C during daytime and nighttime, respectively, representing increases of 0.12 °C and 0.44 °C compared to 2005. For the newly developed UHI areas with UHII exceeding 2 °C, infilling area exhibited greater daytime population heat exposure risk, whereas leapfrogging area showed higher nighttime risk. This study established a connection between UHI spatial expansion and population exposure risk, offering the insights for mitigating urban environmental hazards in the future.