Universities play a role in producing quality resources from their graduate students, so that the quality and accreditation of tertiary institutions are things that need attention. One indicator of higher education accreditation is student graduation on time, so student graduation must be an important concern for tertiary institutions. Based on the results of the documentation, the percentage of students graduating on time is lower than students who are not completing their studies on time, therefore it is necessary to analyze the student's study period to overcome the study period that graduates are not on time using the machine learning method with the Naïve Bayes Classifier algorithm to predict student study period. The research method uses the Naïve Bayes Classifier algorithm method which is part of Artificial Intelligence (AI), which consists of preprocessing, input, process and output. because this method has high accuracy and can work better in real-world cases. The results of predicting the timeliness of student study time at college X with 3553 data using the Naïve Bayes Classifier algorithm method, using WEKA tools succeeded in predicting student study time with 70% data taining and 30% as random testing data with the system. Using 11 attributes, namely study program, GPA, mother's occupation, mother's income, entry period, father's occupation, father's income, route of entry, gender, and school of origin, obtained a percentage of precision value of 54.545%, recall value of 67.220%, and the accuracy level reaches 79.925% which is categorized as good, using the ROC curve calculation to form almost close to (0.1) with an AUC value of 0.844 which is categorized as very good. Based on the results of the percentage accuracy rate, ROC curve and AUC value, the Naïve Bayes Classifier predicts student graduation in the "Good" category