Abstract:Objective To predict the incidence rate of measles by using exponential smoothing method, and to provide a basis for developing measles prevention and control strategies. Methods The data about the incidence rate of confirmed measles cases in a district from 2005 to 2015 were collected. Exponential smoothing prediction was performed by using SPSS18.0 statistical software, and then the prediction results were evaluated. Results Holt-Winters multiplicative model was suitable for forecasting the monthly incidence rate of measles (RMSE=7.69, adj R2=0.52, R2=0.85, Ljung-Box Q=21.91). As for Holt linear trend model, Alpha and P values were 1.000 and 0.000 respectively. Forecasts by the model showed that the level of measles incidence in 2015 was relatively low, and the monthly incidence rates of measles exhibited fluctuations after going through an increase trend. The monthly actual incidence rates of measles displayed low-level fluctuations after undergoing an increase trend and a steep decline trend. Conclusions Forecasts by exponential smoothing model reveal that the relative error for the monthly predicted and actual incidence rates in January-May is only 10.60%, indicating that exponential smoothing method is adapted to predicting the short-term incidence of measles.