Abstract: Objective To explore the application of ARIMA model in the prediction of the incidence of mumps and verify the feasibility and applicability of the model.Methods Incidence data of mumps in Nanjing from January 2004 to December 2012 were fitted with ARIMA model using SPSS 18.0 statistical software. The incidence rate from January to December 2013 was verified by the established model, and then, monthly incidence rates in 2014 were predicted and evaluated. Results A total of 14 871 mump cases were reported from 2004 to 2013, with average annual incidence rate of 21.78 /100 000. Each year the monthly incidence rates were fluctuated around 1.85/100 000. ARIMA (1，0，0) (2，1，0)12 model was set up as the optimal one. The model residual sequence was white noise. The parameters of the model were statistically significant except the constant term. MAPE was 29.63% and R2, 0.76. Monthly incidence rates from January to December 2013 fitting with the model were all within the 95% confidence interval, consistent with the change trend of the actual incidence, which verified the availability of the model. Predicted by the model, the annual incidence rate of mumps in 2014 was 1.48/100 000 and the peak months of incidence were April (incidence rate: 2.33/100 000), May (2.72/100 000), and June (2.52/100 000). Conclusion ARIMA model can be used to fit mumps incidence trends in time series, and for dynamic analysis and short term prediction of the incidence of mumps.