Using datasets from the NCEP/NCAR Reanalysis, we study conventional observational datasets of snow depth, surface air temperature, and snowfall, and the spatio-temporal characteristics of wintertime snow depth over China. We found that snow depth is out of phase between Sinkiang, Northeast China and the area south of the Yellow River with respect to the leading mode of empirical orthogonal functions. Namely, the positive (negative) anomalies in snow depth over northern Sinkiang and Northeast China are associated with negative (positive) anomalies over the area south of the Yellow River valley. Spatial distributions of snow cover in the leading mode are inversely correlated with Arctic Oscillation (AO) in the past thirty years. During the negative AO phase, a cyclonic circulation with its center located at Lake Baikal is observed on the 500-hPa isobaric surface north of 40°N. Meanwhile, an anticyclonic circulation with its center located in Northwest China is also observed south of 40°N. As a result, North China and South China are controlled by cyclonic and anticyclonic circulation, respectively. In North China, anomalous cyclonic circulation associated with AO can result in excessive snowfall, lower temperatures, and ultimately increased snow depth. Meanwhile, in South China, anticyclonic circulation associated with AO may result in less snowfall, higher temperatures, and decreased snow depth. Our studies show that AO may influence wintertime snow depth over China significantly by affecting snowfall and surface air temperature.