<p>Floods are among the most devastating natural disasters worldwide, posing severe threats to human society and economic development. Conducting a comprehensive and scientifically grounded flood risk assessment is therefore essential for effective disaster prevention and mitigation. In this context, the rational definition of disaster-bearing elements and the establishment of a multivariate evaluation framework are key to improving assessment accuracy. This study focuses on the middle reaches of the Yellow River, where a multivariate joint distribution model of flood risk was constructed based on the Copula function, integrating human society and ecosystem as disaster-bearing bodies. Following the IPCC framework of “Risk = Hazard × Exposure × Vulnerability”, soil erosion was introduced as an indicator of ecosystem vulnerability to explore the spatiotemporal patterns of flood risk. The results reveal that the Fenhe, Qinhe, Jinghe, Weihe, and Yiluo River basins represent high-risk flood areas, while the central, western, and northern parts of the study area exhibit relatively lower flood risks. The Beiluo, Jinghe, and Fenhe River basins are more prone to extreme events where both flood peak and volume exceed design thresholds, indicating higher flood risk. Flood risk in the Jinghe, Weihe, and Fenhe River basins is primarily driven by population density and construction levels, whereas the influence of ecosystem factors such as vegetation cover and soil erosion is relatively weaker. This study integrates human-society and ecosystem disaster-bearing elements within a unified analytical framework, enriching the theoretical and methodological foundation of flood risk assessment and providing a scientific basis for regional flood prevention and mitigation planning in the Yellow River Basin.</p>