What is the typical pattern of the southern Yellow Sea circulation in winter? We have attempted to understand the circulation by analyzing CTD data and satellite-tracked drifter positions collected in February 2001 and 2003. As usual in winter, the thermohaline front was developed in a tongue shape. The northwestern part of the front appeared to be partly collapsed in February 2001, suggesting a northwest intrusion of saline water to the Chinese coast, while the Chinese coastal water appeared to flow to the southeast along a 50 m isobath. A fast along-front current was directed to the west in the northern frontal zone. The north-south fluctuations of the drifters are highly correlated each other, although the drifters were distanced and they are also highly correlated with the wind.