Numerous studies have developed new methods for skillful long-lead seasonal streamflow forecasts, especially in the western United States, and most assume that the streamflow forecast skills translate into improved water resources decision making, but there has been little comprehensive demonstration of this. This translation is not straightforward because the decision system is nonlinear. This paper develops a prototype decision support system (DSS) to systematically evaluate the translation of streamflow forecast skills to water resource operational variables and decision variables in the Gunnison River basin (GRB). The DSS consists of two main modules: (1) a multimodel ensemble streamflow forecast module that uses large-scale climate information and issues forecast ensembles, and (2) a decision support module (developed in the generalized decision tool, RiverWare) that captures the seasonal operations and management of the GRB water resources system. Ensembles of streamflow forecast scenarios drive the decision support module, resulting in an ensemble of various variables pertaining to operations of the reservoirs. The DSS is run with streamflow forecast ensembles for the spring season issued on January 1 and April 1 and also the climatological and observed streamflows. The skills of the various variables, e.g., power generated and end of season reservoir storages, are computed. The results demonstrate that the skills from the streamflow forecasts are well transferred to the operational variables, enabling more effective decision making as early as January, when the snow information is partial at best. This integrated and comprehensive prototype DSS is believed to be one of the early attempts to demonstrate the utility of streamflow forecasts in a decision-making context. This DSS framework can be easily transferred to other basins.