SUMMARY
The Selection Of Best Model and Forecasting The Number Of Foreign Tourists Arrival At Bali In 2014. The number of foreign tourists visiting Bali according to figures Central Bureau of Statistics (BPS) recorded the highest in Indonesia, so it is interesting to observe, more specifically related to the prediction of the number of foreign tourists who will visit Bali at some future period. Forecasting method requires stationary data, there are some cases (intervention). In this case the Bali bombings make data on the number of tourist becomes stationary so it needs to be analyzed by comparing the existing forecasting methods. Two methods are applied in this study, which are: Intervention and ARIMA models. The models allow to solve the case of time series data that often not stationary. With the Box-Jenkins procedure and visually observing and forecasting the estimated results of these two methods. The results showed that the best model for the case that the number of tourist arrivals to Bali through Ngurah Rai airport is ARIMA model, which has the smallest value of AIC and RMSE. Keywords: Bali bombing, ARIMA, intervention