Abstract: Objective The study was conducted to determine the performance of the moving average method in early warning of infectious diseases with data of Network Surveillance System for infectious diseases. Methods Data of Network Surveillance System for infectious diseases in 2004, 2005 were collected for calculating the incidence baseline to determine the probability value and . The threshold of early warning was calculated by the equation:W=+Si, with which a control chart of early warning was drawn. The incidence data in 2006 were introduced for interpretation and reaction. Results The probability value was defined as 0.01 on the basis of the harmfulness, severity and controllablility of infectious diseases, then (unilateral)2. The thresholds of weekly early warning were then calculated. A vice epidemic trend could be seen at week 1 and week 2 from the control chart. Conclusion The moving average method for early warning is applicable for high-incidence infectious diseases. The baseline data and probability values of early warning are the key to successful warning.