medical research community is becoming more and more interested in the prevention and prognosis of diabetes mellitus. Objective: This work determined to predict the onset of TYPE_2 diabetes mellitus using patient’s data gathered from several Northern Nigerian healthcare facilities. Method: We applied six different classification algorithms, including Naive Bayes (NB), Random Forest (RF), Neural Network (NN), Linear Discriminant Analysis (LDA), K-Nearest Neighbor (KNN), and Support Vector Machine (SVM). The models’ predictive abilities were then tested using four different metrics. Results: Results of the study show that all six algorithms have outstanding performance. The RF method provided higher accuracy than NN, SVM, LDA, NB and KNN. The accuracy, sensitivity, specificity and kappa of RF were 99.6%, 100%, 99.4% and 99.1%, respectively. Conclusions: We conclude that when compared to the other five models, RF has been shown to provide the highest level of accuracy.