Physically based distributed hydrological models are an invaluable tool for planning and management of water resource projects. However, a reliable prediction from hydrological models can only be expected if their unknown model parameters are estimated accurately. In place of laborious manual calibration of the parameters of the hydrological model, this study presents an automatic calibration scheme for the VIC-RAPID hydrological model using the self-adaptive differential evolution (SaDE) algorithm. The SaDE eliminates the laborious manual tuning of the two control parameters (mutation factor and crossover rate) of the conventional DE algorithm. The proposed approach is demonstrated with a case study in the upper Krishna river sub-basin for estimation of the 15 VIC-RAPID model parameters. The efficacy of the proposed calibration technique for hourly streamflow simulation is evaluated by using standard performance measures such as Nash-Sutcliffe Coefficient (NSE), Coefficient of Correlation (R2), and Percent Bias (PBIAS). Results from this study revealed the potential of SaDE for the parameter estimation of complex hydrological models.