Presented at the 2016 AGU Fall MeetingMoisture transport from the excess of evaporation over precipitation in the global ocean drives terrestrial precipitation patterns. Sea surface salinity (SSS) is sensitive to changes in ocean evaporation and precipitation, and therefore, to changes in the global water cycle. We use the Met Office Hadley Centre EN4.2.0 SSS dataset to search for teleconnections between autumn-lead seasonal salinity signals and winter precipitation over the western United States. NOAA CPC Unified observational US precipitation in winter months is extracted from bounding boxes over the northwest and southwest and averaged. Lead autumn SON SSS in ocean areas that are relatively highly correlated with winter DJF terrestrial precipitation are filtered by a size threshold and treated as individual predictors. After removing linear trends from the response and explanatory variables and accounting for multiple collinearity, we use best subsets regression and the Bayesian information criterion (BIC) to objectively select the best model to predict terrestrial precipitation using SSS and SST predictors. The combination of autumn SSS and SST predictors can skillfully predict western US winter terrestrial precipitation (R2 = 0.51 for the US Northwest and R2 = 0.7 for the US Southwest). In both cases, SSS is a better predictor than SST. Thus, incorporating SSS can greatly enhance the accuracy of existing precipitation prediction frameworks that use SST-based climate indices and by extension improve watershed management.