[Objective] Flood disasters are one of the most significant natural disasters in the middle reaches of the Yangtze River. Measuring the flood disaster resilience levels of the region and exploring the driving factors of their spatial heterogeneity are crucial for enhancing regional disaster resistance and achieving sustainable quality development.[Methods] This paper employs the CRITIC-Entropy Weight Combined Weighting Method and the Geographical Detector, among other methods, to construct a comprehensive flood disaster resilience evaluation index system from four dimensions: society, economy, infrastructure, and environment. It analyzes the spatiotemporal evolution and spatial heterogeneity driving factors of flood disaster resilience in the middle reaches of the Yangtze River from 2012 to 2022. [Results] (1) During the study period, the flood resilience in the middle reaches of the Yangtze River fluctuated and increased from 0.2091 to 0.2629, with only a slight decline in 2020. The structure of flood resilience evolved from “environment–society–infrastructure–economy” to “environment–society–economy–infrastructure”.(2) There are significant differences in flood resilience within the region. 95.24% of the areas are characterized by fluctuating increases. Ganzhou and Ji'an are regions with sustained growth. The differences in social resilience within the region are gradually narrowing. Except for Enshi and Xiangxi, the economic resilience levels of other regions have improved. The infrastructure resilience levels of 25 cities and prefectures have been enhanced. The environmental resilience in the northwestern part of the region is better than that in other areas.(3) The results of factor detection show that the main factors influencing the spatial differentiation of flood resilience in 2012 were population density, terrain relief, and slope. By 2022, these factors had shifted to the number of people receiving minimum living security, the number of large-scale industrial enterprises, and the end-of-year highway mileage.(4) The results of interaction factor detection show that the most influential interaction factor in 2012 was population density ∩ proportion of the tertiary industry. By 2022, it had evolved to per capita GDP ∩ end-of-year highway mileage. [Conclusion] The research results can provide references for urban flood control and disaster relief policies.