The stability of the reservoir bank slope is very important for the service safety of water conservancy projects, especially for high dam projects. When the instability failure occurs, not only will the integrity of the project be damaged but also the safety of the local lives and properties will be greatly threatened. So it is necessary to analyse the stability of reservoir bank slopes. Due to the disadvantages of conventional slope instability analysis methods, including the difficulty of ascertaining the values of rock and soil mechanical parameters and insufficient consideration of dynamic change characteristics, a bank slope instability simulation method based on cellular automata model was proposed with the consideration of the physical and mechanical properties of the slope. With this method, the development path of slope instability was identified dynamically driven by the displacement monitoring data. The results show that the instability failure of reservoir bank slopes is a progressive development process from gradual change to abrupt change, and the instability area is mainly concentrated in the slip accumulation area distributed along the boundary of the strata. This method is applicable to the instability simulation of reservoir bank slopes with complex characteristics and mixed components.