It is very necessary and meaningful to carry out research aiming at traffic signal control sub-area method. However, the accuracy and robustness of traffic signal control sub-area still can not meet the need of practice. First, this paper researches the popular methods and their key techniques. Second, the paper considered following four influencing factors: distance between intersections, major road traffic volume, major road traffic capacity, signal cycle, and was trying to propose a new traffic signal control sub-area method based on fuzzy C-means clustering. Finally, the approach is illustrated by a case study in vissim. The results show this method is satisfactory and wide applicability. Compared with traditional method, major road delay decreases by 30%, and travel time decreases by 20%.