<p>Flash floods pose significant challenges to the stable development of human society, highlighting the need for effective assessment and management of flash flood susceptibility (FFS). This research aims to explore the influence of geomorphological features on FFS using sub-basins as evaluation units, which provide scientific support for accurate flash flood early warning based on disaster monitoring and planning. Firstly, the Dali River Basin was chosen as the study area to simulate the flood processes under different rainfall scenarios using the HEC-HMS hydrological model. Then, the results of the flash flood occurrence under a 40mm-1h rainfall scenario were used to analyze the correlation between basin geomorphological characteristics and FFS, employing only-one-variable Logistic Regression (LR). Additionally, the Least Absolute Shrinkage Selection Operator (LASSO) was employed to select the relevant basin parameters. Subsequently, an FFS assessment model based on selected parameters was developed using LR. This study revealed a significant correlation between the basin shape and the drainage network with FFS, indicating that basins with an equidimensional shape and a well-developed drainage network are more prone to flash floods. The FFS assessment model constructed using geomorphological parameters achieved an Area Under the Curve (AUC) of 0.917 in the Dali River Basin, which can be effectively utilized for assessing FFS in the Dali River and other hydrologically similar basins.</p> Graphical Abstract <p></p>