The National Weather Service (NWS) has initiated an effort to verify the river stage forecasts issued by the NWS River Forecast Centers. As this effort gains momentum the difficulties associated with evaluating river forecasts at a single point or as an aggregate have become more clearly defined. As a first step toward developing a coherent set of statistics that effectively capture the quality of a sample of forecasts, the NWS has begun collecting data and computing a variety of statistics on that data. Simple statistics such as Maximum Error, Root Mean Square Error, Average Error and the like have been found to be effective in some circumstances. While the more traditional meteorological measures of False Alarm Rate and Probability of Detection have been found to be useful in other circumstances. Use of the more sophisticated Distributions Based approach and transforms such as the Normal Quantile Transform have also been investigated.