The research was aimed to evaluate the response of SWAT and HBV models to estimate stream flow in upper Blue Nile basin, Ethiopia. The evaluation was carried out through simulation of runoff in Blue Nile basin, particulary,Gilgel Abay (1653. 64km)2,Angar (4545. 59km2 ), and Maybar (1. 24km2 ) watersheds. The same calibration and validation period was applied for each model for each watershed of respective historical records. During calibration and validation procedure of each model for each watershed, the objective function which indicates the optimized ranges of model parameters that give good model response result were determined by NSE. For HBV model:(NSE=0.81, NSE=0.48 and NSE =0.46) for Gilgel Abay,Angar and Maybar watersheds respectively, SWAT model (NSE=0.81, NSE=0.63, and NSE=0.61) respectively for Gilgel Abay,Angar, and Maybar watersheds. These optimized parameters were validated using independent historical data for each model. The result of SWAT model revaled that efficiency is almost equals to calibration period for Gilgel Abay watershed. The NSE during validation was 0.80, 0.55, and 0.54 respectively for Gilgel Abay, Angar and Maybar watersheds. Using the same data step as SWAT model, HBV model was also validated and better statistics result was obtained in Gilgel Abay watershed too. The estimated value of HBV during validation period was 0.63, 0.45, and 0.34 respectively for Gilgel Abay, Angar and Maybar watersheds. The statistics output hydrograph indicates that SWAT model under estimate the stream flow during wet and dry months for all watersheds during calibration and validation period. Except,Angar watershed with an alternative over and under estimation hydrograph during calibration and validation. In Gilgel Abay watershed in HBV model during calibration period the hydrograph statistics indicates that under estimation of wet season and dry season and over estimation to wet season and there was alternative under and over estimation of flow during validation preiod. Generally, the calibrated and validated hydrograph indicates that the responses of models to extremely wet and dry conditions were not satisfactory. Therefore,further model refinements for extreme events, including the use of season-based model evaluations that could improve process representation of rainfall and runoff events is recommended.