From 2010 to 2022, crime in Indonesia, especially Maluku Province, tends to increase compared to previous years. Considering these problems, a crime rate prediction system is needed so that the Maluku Provincial Police is able to estimate the quantity and type of crime that is likely to occur in the future. One of the prediction methods that has been used for crime prediction is Exponential Smoothing (ES). The Smoothing method is applied to obtain predictions based on time-series data. In this discussion, the author will compare the forecasting methods of Double Exponential Smoothing, and Triple Exponential Smoothing. The Double Exponential Smoothing method is suitable to be used to provide forecasting results when a data has a certain trend pattern. This Triple Exponential Smoothing method is used when there are still dominant expression elements &; seasonal conduite shown in the data. The MAPE value for the Double Exponential Smoothing method is 20.69552 and for the Triple Exponential Smoothing method is 30.48323, it can be said that the MAPE value of the Double Exponential Smoothing method is smaller than the Triple Exponential Smoothing method. So that the Double Exponential Smoothing method is more accurate than the Triple Exponential Smoothing method to predict the crime rate.