The snow cover over the Tibetan Plateau (TP), as a sensitive indicator of climate change, has a significant impact on regional and even global climate. MODIS 8- day snow cover extent products and ERA-Interim reanalysis data were employed to study the spatial and temporal variability of the snow cover over the TP and its relation to Arctic Oscillation (AO) by climatological statistical diagnosis. The spatial distribution of winter snow cover over the TP is far from uniformity, with high snow cover fractions (SCF) at the western edge and the southeast part of the TP but scarce snow in the northern and central parts. It is found that the SCF is out of phase between the eastern and western parts of the TP with respect to the leading mode of empirical orthogonal functions (EOF1), namely, the positive (negative) anomalies in SCF over the eastern part of the TP are associated with negative (positive) anomalies in SCF over the western part. This pattern is positively correlated with AO. During the positive AO phase, the East Asian Trough weakens, together with intensive Southern Branch Trough. The warm moist flows easily lift to the eastern part of the TP because of intensive Subtropical High over the Western Pacific and result in excessive snowfall, while an anomalous anticyclone with its center to the southwest of the plateau leads to sinking dry air flows over the western TP, which is not prone to snowfall, and the corresponding higher surface temperature is also against maintaining the snow cover. During the negative AO phase, the East Asian Trough strengthens and so does the East Asian winter monsoon, with dry cold air flows over the eastern part of the TP, leading to less snowfall. On the other hand, an anomalous cyclone centered to the southwest of the plateau makes it easier for the warm moist flows from the Bay of Bengal and Arabian Sea to lift to the western part of the TP and meet the cold air from Siberia, thus prompting more snowfall over the western part of the TP.