Flood hazard due to dam failure is one of the most catastrophic events that could occur worldwide. The dam failure certainly damages the properties downstream, but loss of life and the size of the population at risk dramatically vary with the hazard level due to dam failure and the available warning time. The hazard level depends on the outflow volume and its temporal variation, which are significantly affected by breach geometry and opening dimensions. Therefore, simulation of dam breach events and the resulting floods are very essential to characterize and reduce hazards associated with potential dam failures. As the major portion of dams worldwide is of embankment type, and due to the availability of data, this paper is devoted to improve the estimation of embankment dam breach parameters. A new set of empirical equations to estimate average breach width, depth to final breach elevation, and breach formation time is generated using the data from 166 failed embankment dams. Shuffle Complex Evolution Metropolis is used for parameter estimation while resampling using bootstrapping is used to assess the uncertainty in the generated equations. It is shown that the new set of equations gives better results than the existing ones.