Snow-water equivalent (SWE) data measured at several hundred montane sites in the western United States are used to examine the historic effects of El Nino and La Nina events on seasonal snowpack evolution in the major subbasins in the Columbia and Colorado River systems. Results are used to predict annual runoff. In the Columbia River Basin, there is a general tendency for decreased SWE during El Nino years and increased SWE in La Nina years. However, the SWE anomalies for El Nino years are much less pronounced. This occurs in part because midlatitude circulation anomalies in El Nino years are located 35¿ east of those in La Nina years. This eastward shift is most evident in midwinter, at which time, SWE anomalies associated with El Nino are actually positive in coastal regions of the Columbia River Basin. In the Colorado River Basin, mean anomalies in SWE and annual runoff during El Nino years depict a transition between drier-than-average conditions in the north, and wetter-than-average conditions in the southwest. Associations during La Nina years are generally opposite those in El Nino years. SWE anomalies tend to be more pronounced in spring in the Lower Colorado River Basin. Our predictions of runoff reveal modest skill for scenarios using only historic El Nino and La Nina information. Predictions based on the water stored in the seasonal snowpack are, in almost all cases, much higher than those based on El Nino-Southern Oscillation (ENSO) information alone. However, combining observed midwinter snow conditions with information on seasonal snowpack evolution associated with ENSO improves predictions for basins in which ENSO signals exhibit strong seasonality. ¿ 2001 American Geophysical Union