Atmospheric rivers (ARs) account for a large portion of total precipitation in the western United States (WUS), specifically in winter months. Interannual variability of precipitation is also high in this region due to large variability in landfalling ARs. Accurate prediction of ARs and precipitation remains challenging due to large noise associated with internal variability, the nonlinear response of ARs to variability in sea surface temperature and atmospheric circulation, and relatively short observational records. To improve the prediction of landfalling ARs and precipitation, we examine the relationships between two prominent climate modes, namely the El Niño/Southern Oscillation (ENSO) and Madden–Julian oscillation (MJO), and winter ARs and associated precipitation (AR precipitation) by using observations and a large ensemble of Weather Research and Forecasting (WRF) model hindcasts (27 km resolution) over the period 1981–2017. We found that WRF reproduces the spatial and temporal patterns of AR frequency, including the climatological mean and interannual variability. WRF is also capable of reproducing the mean and variability of seasonal total precipitation in WUS and captures the variability of AR precipitation, despite an underestimation of AR precipitation. Composite analysis shows that ENSO events have distinct effects on AR activity at local, regional, and basin scales. Some MJO phases can lead to anomalously high/low AR activity, depending on the phase of MJO, time lag, and subregion of WUS. We also found that ENSO substantially modulates the MJO-AR relationship. Depending on the phase of ENSO, the MJO-AR relationship can produce different anomalies in AR frequency. If we compare El Niño-MJO and La Niña-MJO periods, only ~50% of their anomalies in AR frequency have the same sign (i.e. both positive or negative). A parallel analysis for AR precipitation further confirms the joint influence of ENSO-MJO on AR precipitation. Our findings highlight the necessity of evaluating concurrent effects of different climate modes on ARs and precipitation, and they may shed light on a path toward accurate subseasonal-to-seasonal prediction of ARs and precipitation over WUS.