Abstract: Under global climate change, extreme events such as heatwaves and droughts tend to occur more frequently and concurrently. To improve the understanding of the occurrence patterns and formation mechanisms of the compound events, this study employed a distributed hydrological model to simulate runoff sequences in the Yangtze and Yellow River basins during the historical period of 1971—2020. The standardized runoff index was calculated to identify hydrological droughts, while a threshold method was applied to detect heatwave events. A compound hydrological drought-heatwave index was constructed to analyze the spatial distribution characteristics of such compound events in terms of duration, frequency, and intensity. The causes were investigated from perspectives including moisture budget, land-atmosphere energy transfer, and atmospheric circulation anomalies. The results show that hydrological droughts in both basins lasted 2—6 months, occurring 0.1—3 times annually with intensities ranging from 0.1 to 8; heatwaves lasted 4—6 days, occurring 1.6—2.6 times annually with intensities of 6—12; and the compound events lasted 4—7 days, occurring approximately 0.1—2 times annually with intensities of 6—14. In the early phase of the compound events, heat accumulation was dominated by diabatic heating, while in the mid-to-late phase, the negative feedback effect of temperature advection and adiabatic heating becomes significantly enhanced. During the events, both longwave and shortwave radiation anomalies increased in the two basins, intensifying land-atmosphere feedback. In addition, strengthened South Asian and Western Pacific subtropical high-pressure systems reduced precipitation and impeded moisture transport, leading to runoff reduction of 7.0% in the Yellow River basin and 8.6% in the Yangtze River basin. The imbalance in water vapor budgets and enhanced land-atmosphere feedback are identified as the core mechanisms driving the formation of the compound events.