Agriculture forms the foundation of Indian economy. It provides employment to millions of people worldwide. The ongoing challenge for Indian farmers is that they often choose crops without considering environmental factors, leading to major productivity losses. We performed predictions using three algorithms, random forest, gradient boost classifier and logistic regression on Agricultural Crop Yield in Indian States Dataset. By reviewing the performance indicators of different ML models, we have come into the conclusion that RF algorithm delivers the best results with the least value of MSE & MAE, and the highest R2 score, which is an indicative of better accuracy and predictive power.