This study investigates the dominant modes of interannual variability of precipitation in dry (October-April) and wet (May-September) seasons over southwestern China through empirical orthogonal function (EOF) analysis of monthly precipitation reported by 26 meteorological stations during 1980-2009. The dominant modes are shown to be specific in each season. In the dry season, the first leading mode is generally distributed uniformly throughout the region with quasi-biennial oscillation, and the second leading mode is characterized by an obvious two-three year period of southeast and northwest opposite distribution from the mid-1990s until recently. In the wet season, the first leading mode shows the same distribution with an obvious two-four year period in the early 1990s, and the second leading mode shows spatial distribution of a meridional dipole pattern during a significant four-year period. The third leading mode is of a zonal dipole distribution and exhibits changes in two-four-year periods. We further examine the relationships of dominant modes of dry-and wet-season precipitation with large-scale atmospheric circulation and sea surface temperature (SST). In the dry season, the first leading mode is closely related to Arctic Oscillation (AO). For this mode, high-and mid-latitude opposite distribution in the 500-hPa height field is detected in addition to a meridional dipole-like structure in the North Atlantic and North Pacific SST fields. The second leading mode is related to a meridional dipole-like structure in the 500-hPa height field over Eurasian Continent and SST anomalies over the mid-latitude North Pacific. In the wet season, the first leading mode is significantly and negatively correlated with North Atlantic Oscillation (NAO). This mode is related to a dipole pattern between high and middle latitudes in the 500-hPa height field and a meridional tripole-like structure in the North Atlantic SST field. The second leading mode is associated with a meridional tripole pattern in the 500-hPa height field over the Eurasian Continent and a nearly coincident distribution of North Pacific SST anomalies. The third leading mode is related to the El Ni?o Modoki and South Asia high. The zonal tripole-like structure in the equatorial Pacific SST field may act as a potential predictor of the third mode.