Abstract — Real Analysis is a course that emphasizes verification and skills in analyzing statements, so it becomes difficult for most students. With the online lecture system, the success of learning in this course will also has an impact on students. The purpose of this study was to determine the shape of the model, the factors that influence the student graduation, and the factors that have the greatest chance of graduating a student of the Mathematics Department of Mathematics and Natural Sciences in UNP in studying the Real Analysis course during online lecture using probit regression analysis. The data analysis used was the estimation of probit regression parameters, the significance test of the model using the G test, the significance test of the parameters using the Wald test, selecting the best model, interpreting the model and calculating the probability of passing students from the best model. Based on the research of the independent variables that affect the graduation of Mathematics Department students of Mathematics and Natural Sciences in UNP in studying Real Analysis courses during the implementation of online lecture is the background factor. Keywords — Probit Regression Analysis, Real Analysis, Student Graduation