It is necessary to understand the hydrological processes undergoing at small segments in order to apprehend the complex large river systems and this can be achieved by application of hydrological models. This study aims to check the applicability of soil and water assessment tool (SWAT) model over the Kharun River Basin, Chhattisgarh, India. The limitation of data at proper scales is the major problem that hinders research on this area. The meteorological data were collected from climate forecasting system reanalysis (CFSR) database for 1994–2014. The SWAT simulations generated runoff, which was auto-calibrated with the observed values through SWAT-CUP. Generalized uncertainty likelihood estimation (GLUE) optimization approach was used for sensitivity analysis of various parameters during auto-calibration. The results reveal for a moderate correlation between the observed and modelled values at daily time scale i.e. the Nash-Sutcliffe efficiency and R2 values are close to 0.5. This highlights the fact that model application needs to be carried out carefully with processed and reliable data sets. Also, some of the popular models may as well be not suitable for few basins with multiple complexities. The possible reasons for such moderate performance of the model are also discussed. This work will be helpful to the decision makers for better planning and management of the basin.