Online discussions are popular in virtual classrooms since they provide students with opportunities for more effective interactions than possible in traditional classrooms. However, grading online discussions is well-known to be a time-consuming task. In this paper, we propose a method for automating the partial grading of online discussions using social network analysis techniques. In particular, we propose to use a centrality index, which reflects the number of responses to a student’s messages as a proxy measure to evaluate the quality of students’ posts and hence determine participation levels in discussions. The method was tested for grading discussion board assignments in a virtual class from an online graduate school. Although, the method was found to perform quite well in a range of situations, its use is advocated mainly as a grading aid because in certain conditions inaccurate grades may result. Furthermore, use of the tool can also be envisaged for improving performance levels in online discussions because of its potential as facilitation and motivational aids.