A substantial body of discharge rate data has been developed over the past half century applicable for validation of single and two-phase discharge models. This paper applies a wide range of test cases and compares predictions with test data for two types of discharge model: (a) the energy balance model, and (b) the non-equilibrium model of Diener and Schmidt. The latter enhances the original homogeneous equilibrium model of Leung. This exercise reveals possible inconsistency between experimental datasets as much as it provides confirmation of the accuracy of the models, but both models are shown to provide adequate predictions within a factor of two and generally better. (C) 2009 Elsevier B.V. All rights reserved.