The integrated prediction decision making for typhoon tracks by using the Multi Criteria Evaluation (MCE) method is made. It is shown that the MCE can provide a reasonable integrated prediction for typhoon tracks. An objective integrated decision for typhoon track forecasting, which is a neural network to simulate the human being, is made instead of the traditional experiment one. Parallel analyzing results by using typhoon track forecasting data, 10 year′s special landing typhoons, the full year for 2000 and typhoon 9711 (Winnie ) with abnormal track indicated that MCE method gives a good behaviour.