Introduction: Swine Flu cases are rarely observed but are to be found subsequently increasing in many countries and also in India. They can be treated and diagnosed if it is detected at an early stage. However, many times it is difficult to predict as the symptoms are almost similar to other viral fevers. Many researchers used various algorithms and different approaches to predict swine flu. Aim and Objective: In this paper, we propose Naive Bayesian classifier for an efficient prediction, by comparing the symptoms and applying significant signs and prediction factors to Bayes theorem to predict the patient__ampersandsignrsquo;s condition. Method: Here in this work, the machine learning algorithm is used on the available data to reduce the time and to predict the disease perfectly. We collected 14 sample records of different swine flu suspected patients from King George Hospital, Visakhapatnam, India, based on the sample reports. We prepare the training SF (Swine Flu) dataset. The ESFP (Efficient Swine Flu Prediction) is first trained by using SF (Swine Flu) dataset with some of the significant symptoms observed in Swine Flu cases. Conclusion: The first signs and the prediction values are applied to Naive Bayesian Classifier by using the training set, and then the test set is applied to predict the disease prevailing or not..