The majority of the public prefer government clinics for health treatment. This preference results in a high number of outpatients at the government clinics and is the leading factor in the long waiting time for patients. The waiting time is the period that a patient has to wait at the clinic; from the registration process until consultation with the medical officer. Long waiting time is the leading cause of patient dissatisfaction compared to other factors. The dissatisfaction is especially so when there is no convenient waiting facility. The objective of this study is to analyse outpatient visits patterns to government outpatient clinics. The recognised patterns will form the basis to recommend which government outpatient clinics are the best to visit on a specific time to minimise waiting time. Then, the waiting time is predicted based on the arrival time. Fifteen government outpatient clinics in the state of Selangor that uses an electronic medical record system provided the data for this study. From the data analysis, there is a correlation between patient waiting time, day, the month of visit, patient age and consultation time. The waiting time is predicted using regression and classification machine learning method. The best regression model is produced by Linear Regression algorithm with mean squared error (MSE) -0.158882 and the best classification model is created by Linear Discriminant Analysis (LDA) with 98.91% accuracy and Support Vector Machine (SVM) with 95.90% accuracy.