Utilizing advanced predictive modelling techniques, the "Smart Health Prediction Using Machine Learning" system serves as a dynamic platform for forecasting the health conditions of patients or users based on the symptoms they input. The application accommodates three distinct user roles: user/patient, doctor, and admin. Upon user input of symptoms, the system employs a sophisticated algorithm to evaluate and predict the likelihood of specific diseases. At the core of this intelligent health prediction system lies the Naive Bayes Classifier, a machine learning model that, during its training phase, incorporates a comprehensive array of features to calculate the probability percentage associated with different diseases. The Naive Bayes Classifier, having assimilated a diverse set of features during training, demonstrates its ability to make accurate predictions regarding the likelihood of diseases. The output of this classifier provides users and patients with valuable insights into their health conditions, contributing to early disease detection and offering a clear comprehension of the prevailing medical circumstances. The multifaceted user access, encompassing patient, doctor, and admin roles, adds a layer of versatility to the application, catering to the distinct needs and perspectives of various stakeholders. This innovative approach to health prediction not only underscores the potential of machine learning in healthcare but also emphasizes the importance of early intervention and informed decision-making for individuals managing their health.