Based on the large-scale model tests, a simplified dam breach model for homogeneous cohesive dam due to overtopping failure is put forward. The model considers headcut erosion as one of the key homogeneous cohesive dam breaching mechanisms and we calculate the time-averaged headcut migration rate using an energy-based empirical formula. A numerical method is adopted to determine the initial scour position at the downstream slope in terms of the water head and dam height, and the broad-crested weir equation is utilized to simulate the breach flow. The limit equilibrium method is used to analyze the stability of breach slope during the breach process. An iterative method is developed to simulate the coupling process of soil and water at each time step. The calculated results of three dam breach cases testify the reasonability of the model, and the sensitivity studies of soil erodibility show that sensitivity is dependent on each test case’s soil conditions. In addition, three typical dam breach models, NWS BREACH, WinDAM B, and HR BREACH, are also chosen to compare with the proposed model. It is found that NWS BREACH may have large errors for cohesive dams, since it uses a noncohesive sediment transport model and does notconsider headcut erosion, WinDAM B and HR BREACH consider headcut erosion as the breaching mechanism and handle well homogeneous cohesive dam overtopping failure, but overall, the proposed model has the best performance. Based on the large-scale model tests, a simplified dam breach model for homogeneous cohesive dam due to overtopping failure is put forward. The model considers headcut erosion as one of the key homogeneous cohesive dam breaching mechanisms and we calculate the time-averaged headcut migration rate using an energy-based empirical formula. A numerical method is adopted to determine the initial scour position at the downstream slope in terms of the water head and dam height, and the broad-crested weir equation is utilized to simulate the breach flow. limit equilibrium method is used to analyze the stability of breach slope during the breach process. An iterative method is developed to simulate the coupling process of soil and water at each time step. The calculated results of three dam breach cases testify the reasonability of the model , and the sensitivity studies of soil erodibility show that sensitivity is dependent on each test case’s soil conditions. In addition, three typical dam breach models, NWS BREACH, WinDAM B, and HR BREACH, are also found to match against the proposed model. It is found that NWS BREACH may have large errors for cohesive dams, since it uses a non-cohesive sediment transport model and does notconsider headcut erosion , WinDAM B and HR BREACH consider headcut erosion as the breaching mechanism and handle well homogeneous cohesive dam overtopping failure, but overall, the proposed model has the best performance.