Dengue Hemorrhagic Fever (DHF) is one of the common and fatal diseases in Indonesia. Jambi city is one of the dengue-endemic areas in Jambi province. To reduce the incidence rate of dengue, an early warning based on forecasting is necessary. Time-series forecasting of DHF can provide useful information to support and help public health officers for planning on DHF prevention. This paper compares two methods for Time-series forecasting of DHF incidence, namely seasonal autoregressive integrated moving average (SARIMA) and Long Short-Term Memory (LSTM). The forecasting performance is assessed using the monthly number of DHF incidence data from January 2012 to April 2019 were acquired from the district health offices. To show the effectiveness of the model, the performances are evaluated based on two metrics: mean absolute error (MAE) and root mean square error (RMSE). In the first analysis, we found that the SARIMA ((1,0,0) (1,0,0)12) is the most suitable model to predict the number of monthly DHF incidence with RMSE value of 30.07 and MAE 18.97, and the second one used the LSTM with one hidden layer (1-64-1) architecture with RMSE of 30.41 and MAE 18.27. Based on the experiment between SARIMA and LSTM perform relatively well to predict the future.