AbstractLand-atmosphere (L-A) feedbacks are important for understanding regional climate functioning. However, the accurate quantification of feedback strength is challenging due to complex, nonlinear interactions and varying background atmospheric conditions. In particular, the role of cloud water in the terrestrial water cycle is often ignored or simplified in previous L-A feedback studies, which overlook the relationship between evapotranspiration (ET) and cloud water (TQC). This study diagnoses the interactions between $$\:ET$$, $$\:TQC$$ and its dynamics ($$\:\varDelta\:TQC/\varDelta\:t$$) under different atmospheric conditions by conducting correlation and a novel scaling analysis, based on a coupled regional climate model simulation. Contrasting correlation relationships between $$\:ET$$, $$\:TQC$$ and $$\:\varDelta\:TQC/\varDelta\:t$$ were identified, indicating the positive feedback between $$\:ET$$ and the dynamics in cloud water. Two types of positive scaling relationships between $$\:ET$$ and $$\:\varDelta\:TQC/\varDelta\:t$$ were identified by K-means clustering. The analysis shows a contrasting north-south distribution of the scaling relationship that is similar to the spatial distribution of energy-limited and water-limited $$\:ET$$ regimes, highlighting the role of ET regimes in modulating the $$\:ET$$ - $$\:\varDelta\:TQC/\varDelta\:t$$ scaling relationships. Moreover, the feedback strength and scaling relationship are affected by atmospheric moisture flux dynamics, providing remote moisture sources and altering dry/wet conditions. Our results highlight the role of cloud water in the atmospheric part of the L-A process chain and reveal the effect of different atmospheric conditions on L-A interactions based on the new analysis framework.