Abstract: The quantitative prediction of the number of monthly new outbreaks of CSF can provide animal disease prevention and control departments the basis for taking selective preventive measures. Grey model and ARMA model were chosen to predict the number of monthly new outbreaks of CSF in Guizhou province, China. By grey model and ARMA model, outbreaks of CSF in Guizhou province were predicted according to the statistical data from veterinary bulletin from 2005 to 2008, and both models were assessed by the statistical data in 2009. Mean absolute error of grey model and ARMA model was 1.84 and 1.48 respectively, while mean absolute percentage error of two models was 0.272 and 0.229. The results showed that the prediction precision of ARMA model was higher than grey model, and ARMA model was feasible and effective for the prediction of CSF outbreaks in Guizhou province, China.