The increasing demand for air travel is stressing the current Air Traffic Control (ATC). This is likely to cause both safety and performance degradation in the near future. In order to solve this problem, increasing the automation level of ATC is an important development direction. So flight behavior recognizing is becoming a key technique for ATC, for it is the basis of other function of ATC, such as landing scheduling, conflict detection, and so on. This paper introduced support vector machine (SVM) to solve flight behavior recognizing in terminal area, and designed multi-classification algorithm flow. The simulation results show that SVM is equal to this task.