The Baijiabao landslide in the Three Gorges Reservoir (TGR) region of China has exhibited episodic movement behavior since the impoundment of the TGR in 2003. This study investigates the causes of this episodic movement using multiple-time scale analysis. Cumulative displacements recorded at four manual monitoring sites from 2007 to 2020 and three automatic monitoring sites from November 2017 to August 2020 are analyzed. Geographical detectors are used to identify key individual and combined predisposing factors at monthly, weekly, and daily scales. The results show that continuous deformation over a dozen days induces episodic movement at the monthly scale. At the short-term scale, reservoir water levels are a more significant factor than rainfall in causing landslide deformation. At the long-term scale, both reservoir levels and rainfall conditions significantly influence episodic movement. Additionally, combined predisposing factors, consisting of one rainfall-related and one reservoir level-related factor, can nonlinearly enhance each other, providing better correlation with landslide deformation.The Baijiabao landslide in the Three Gorges Reservoir (TGR) region of China has exhibited episodic movement behavior since the impoundment of the TGR in 2003. This study investigates the causes of this episodic movement using multiple-time scale analysis. Cumulative displacements recorded at four manual monitoring sites from 2007 to 2020 and three automatic monitoring sites from November 2017 to August 2020 are analyzed. Geographical detectors are used to identify key individual and combined predisposing factors at monthly, weekly, and daily scales. The results show that continuous deformation over a dozen days induces episodic movement at the monthly scale. At the short-term scale, reservoir water levels are a more significant factor than rainfall in causing landslide deformation. At the long-term scale, both reservoir levels and rainfall conditions significantly influence episodic movement. Additionally, combined predisposing factors, consisting of one rainfall-related and one reservoir level-related factor, can nonlinearly enhance each other, providing better correlation with landslide deformation.