Flood Forecasting at Narmada River basin, Madhya Pradesh, India is carried out using an ensemble technique. The presented model has been compared with the individual models such as Support Vector Machine (SVM-EPS & SVM-NU), Random Forest, Multiple Linear Regression, Gradient Boosting, Extreme Gradient Boosting, and various performance measures has been used in order to evaluate and compare the performances of individual models with the stacked model. The evaluation results shows that the stacked model surpass all the individual models and shows improvement in the prediction of peak flows when compared to the individual models with minimal error. © 2020 Ecological Society of India. All rights reserved.