Users of hydrologic predictions need reliable, quantitative forecast information, including estimates of uncertainty, for lead times ranging from less than an hour during flash flooding events to more than a year for long-term water management. To meet this need, operational agencies are developing hydrological ensemble forecast techniques to account for sources of uncertainty such as future precipitation, initial hydrological conditions, and hydrological model limitations including uncertain model parameters. Research advances in areas such as hydrologic modeling, data assimilation, ensemble prediction, and forecast verification need to be incorporated into operational forecasting systems to assure that the state-of-the-art products are reaching the forecast user community. The Hydrologic Ensemble Prediction EXperiment (HEPEX) has been formed to develop and demonstrate new hydrologic forecasting technologies, and to facilitate the implementation of beneficial technologies into the operational environment.