Educational Data Mining is an emerging research area with vast algorithms to analyse hidden patterns in student’s data and discover new knowledge for academic counselling and improvement. Recent deployment of Classification and Clustering algorithms on educational data has proven to be more successful in classifying data instances correctly. One key focus of higher education is to provide quality education and guidance throughout the academic year. This can only be achieved if student’sfailure rate can be averted in final exams by determining underperforming students earlier on in the semester and sampled for academic counselling and recommendation by school authorities. This study presents a Naïve Bayes Classification approach in predicting student’s final grade. The Classifier model was built on the training dataset of previous students who offered the same course with a predictor accuracy rate of 88%. This deployed algorithm will help under-performing students to improve their final exam score.