Heat is identified as a key climate risk in Europe. Vulnerability to heat can be aggravated by enhanced exposure (e.g., urban heat island), individual susceptibility (e.g., age, income), and adaptive capacity (e.g., home ownership, presence of vegetation). We investigated the spatial and temporal patterns of the environmental and social drivers of vulnerability to heat in the Paris region, France, over the 2000-2020 period, and their association with mortality (restricted to 2000-2017). Daily temperatures were modeled for the 5,265 IRIS of the Paris region for 2000-2020. Annual land use and socioeconomic data were collected for each IRIS. They were used to identify a priori five classes of heat-vulnerable areas based on a cluster analysis. The temperature-mortality relationship was investigated using a time-series approach stratified by clusters of vulnerability. The Paris region exhibited a strong urban heat island effect, with a marked shift in temperature distributions after 2015. The clustering suggested that the most heat-vulnerable IRIS in the Paris region have a high or very high exposure to temperature in a highly urbanized environment with little vegetation, but are not systematically associated with social deprivation. A similar J-shape temperature-mortality relationship was observed in the five clusters. Between 2000 and 2017, around 8,000 deaths were attributable to heat, 5,600 of which were observed in the most vulnerable clusters. Vulnerability assessments based on geographical indicators are key tools for urban planners and decision-makers. They complement the knowledge about individual risk factors but should be further evaluated through interdisciplinary collaborations.