The purpose of this study is to investigate the application of emerging knowledge discovery methodologies in analyzing student profiles to predict the performance of a student in a course. Knowledge discovery is the research area concerned with analyzing existing information and extracting implicit, previously unknown, hidden and potentially useful knowledge in an automated manner. The discovered knowledge is often represented by a set of rules or mathematical functions which has practical application. This type of knowledge can enable instructors to accommodate each student's learning needs and abilities as well as aid the students in appropriate course selection. In this paper we present a pilot study which demonstrates the analysis of student profiles from 60 students. The methodology used for knowledge discovery is based on Rough Set Theory which combines theories such as fuzzy sets, evidence theory and statistics. The results of the pilot study show that the knowledge discovery methodologies are likely to discover knowledge which may be overlooked using traditional statistical approaches. Our preliminary results indicate that knowledge discovery methodologies can be successfully used in predicting student performance. Based on the experiences gained from this work, specific future research directions and tasks to ensure a successful comprehensive implementation are discussed.