At the present time, the amount of data in educational database is increasing day by day. These data enclose the valuable concealed information that can lift the student’s performance. Among all classification algorithms, decision tree is most effective algorithm. Decision tree provides the more correct and relevant results which can be beneficial in improvement of learning outcomes of a student. The ID3, C4.5 and CART decision tree algorithms are already implemented on the data of students to anticipate their accomplishment. All three classification algorithm have a limitation that they all are used only for small database. So, for large database we are using a new algorithm i.e. SPRINT which removes all the memory restriction and accuracy problem arrives in other algorithms. It is fast and scalable than others because it can be implemented in both serial and parallel fashion for good data replacement and load balancing. In this paper, we are representing a new SPRINT decision tree algorithm which will be used to solve the problems of classification in educational data system.