©2019. American Geophysical Union. All Rights Reserved.Flood hazard maps are useful tools for land planning and flood risk management in order to increase the safety of flood-prone areas that can be inundated in the event of levee failure. However, flood hazard assessment is affected by various uncertainties, both aleatory and epistemic. The flood hazard analysis should hence take into account the main sources of uncertainty and quantify the confidence of the results for a given design flood event. To this end, this paper presents a probabilistic method for flood hazard mapping, which considers uncertainty due to breach location and failure time. A reliability analysis of the discretized levee system, performed using the concept of fragility function, enables the preselection of a set of levee sections more susceptible to failure. The probabilities of the breach scenarios (characterized by different breach locations and times) are then calculated using the probability multiplication rule, neglecting multiple breaches. The method is applied to a 96-km levee-protected reach in the central portion of the Po River (Northern Italy) and to an adjacent 1,900-km2 flood-prone area on the right-hand side of the river, with a focus on the piping breach mechanism. The numerical simulations are performed through a combined 1D-2D hydrodynamic model using widespread free software. The results show that the method is effective for probabilistic inundation and flood hazard mapping. In addition, it has the advantage of requiring a smaller computational effort in comparison with the methods based on a classic Monte Carlo procedure.