The successful operation of reservoirs for flood control is determined by decisions on the time and volume of the initial release of reservoir storage, which is highly dependent on the inflow prediction. This research intends to assess the reliability of inflow prediction by using the HEC-HMS model generated with rainfall data series from 2011 to 2021 in the Dodokan watershed system, which has Batujai-Pengga series reservoirs. The methodology is to test the model in real-time with input rainfall forecast data based on “The Weather Channel Prediction” for the next two days. The input data was automatically imported into the HEC-HMS model on a scheduled basis using a combination of VBA, Python, and Jython programming languages. However, the model parameters were calibrated by evaluating the similarity of the model output with the Karang Makam AWLR observational data. The result of HEC-HMS modeling calibration is obtained as follows: the NSE is 0.42 due to the bias percent of 9.52%; the RMSE is 0.8, which indicates that the model has well represented the process of converting rainfall into streamflow. In the application of the model by real-time input at inflow in Batujai and Pengga Dam, which is updated every day (24 h), the NSE, percent bias, and RMSE values are obtained, each show that the model can represent the flow condition in the field.