The variability of winter precipitation over the northeastern United States and the corresponding teleconnections with five dominant large-scale modes of climate variability (Atlantic Multidecadal Oscillation, AMO; North Atlantic Oscillation, NAO; Pacific-North American pattern, PNA; Pacific Decadal Oscillation, PDO; and El Niño–Southern Oscillation, ENSO) were systemically analyzed in this study. Three leading patterns of winter precipitation were first generated by empirical orthogonal function (EOF) analysis. The correlation analysis shows that the first pattern is significantly correlated with PNA and PDO, the second pattern is significantly correlated with NAO and AMO, and the third pattern is significantly correlated with ENSO, PNA, and PDO. To verify the physical sense of the EOF patterns and their correlations, composite analysis was applied to the precipitation anomalies, which reproduced the three EOF spatial patterns. Multiple linear regression models generated using indices of all five modes of climate variability show higher explained variances. Composite analyses of geopotential height, sea level pressure, relative humidity, and moisture flux field were performed to find the physical mechanisms behind the teleconnections. When the findings are applied to the extreme drought of the 1960s, it is found that besides a continuous negative NAO pattern, a negative PNA pattern and La Niña conditions also contributed to the drought of winter season by influencing moisture flux and the position of storm tracks. Another case, the 2009/2010 winter with positive precipitation anomalies over the coastal region, is found to be resulted from circulation patterns dominated by major El Niño condition with high-PNA and PDO indices