Analysis on data related to physiological tests of children related to 21 sports and prediction of top 5 sports suitable using various machine learning tools such as Decsion Tree, SVM , Linear Regression , Logistic Regression and KNN. The model achieved 97% accuracy on the data of 70,000 children