Meltwater released by shallow snow covers of the Canadian Prairies is an important water resource to the region. Therefore, many water-management agencies are interested in methods of forecasting streamflow and seasonal water yield from snowmelt. Reliable, accurate forecasts, however, require information of the time of melt and snowmelt rates and volumes. At the present time these quantities are usually estimated by simple temperature-index methods, which have not proven successful in open grassland environments.The paper describes the development and testing of a snowmelt model that uses the energy equation as its physical framework. Empirical procedures for evaluating radiative, convective, advective, and internal-energy terms from standard climatological measurements are presented. Algorithms for accounting for changes in the energy terms in a daily energy-balance model are described.The application of the energy-budget snowmelt model (EBSM) for predicting ablation and simulating streamflow from small and large watersheds is evaluated. It is demonstrated that the EBSM is workable in an operational forecast system and when incorporated within such a system leads to general improvement in synthesizing streamflow from snowmelt.