<p>Effective water resources management in river basins requires an accurate estimation of flood discharge. This study aimed to estimate annual discharge using the Response Surface Model (RSM) in the Darehroud sub-basins of Ardabil Province in Iran and to compare these results with those obtained from various empirical methods. In this study, we reviewed and evaluated several empirical methods for estimating annual discharge, including the Coutagine, Turc, IDOI, ICAR, Lacey, Justin, Inglis, De’Souza, and Khosla methods, specifically in the Darehroud sub-basins. We collected discharge data over 21&#xa0;years from seven hydrometric stations and completed any incomplete data using statistical techniques. Using ArcGIS and WMS software, we determined the physiographic characteristics of the sub-basins, including area, slope, shape factor, average height, time, and curve number. Additionally, we obtained average yield and temperature at the basin level. The results indicated that the determination coefficients (R²) for the empirical methods were as follows: Coutagine (5%), Turc (5%), IDOI (7%), ICAR (15%), Lacey (12%), Justin (2%), Inglis (7%), De’Souza (7%), and Khosla (7%). These low values suggest that none of the empirical methods provide reliable results. In contrast, the RSM predicted determination coefficients of 98% (R²) and 99% (adjusted R²). This high level of accuracy was further confirmed by the adequacy precision index, which was evaluated at 129.2. The analysis showed that the differences between the values predicted by the model and the measured values were not within the 1% confidence level of the mean. In conclusion, this study highlights that the Response Surface Model is sufficiently accurate for calculating annual flood discharge in the sub-basins of the study area.</p>