Maximizing the success rate of students is a central objective of any educationa linstitution. Success rate is correlated to students having better short term memory and knowledge retention ability. Student retention is used as a measure of performance that impacts the success of students, schools, educators, communities, and society. Predicting the likely success rate of students is critical for maximizing this success rate through action-taking. Action-taking allows the educational institutions to provide a counter-measure towards students projected to fall below the passing grade. This project aims to predict the likely success rate of students given past results for the OSCE which is a practical examination method, currently applied on Umeå university students. This study implemented and examined four different supervised learning algorithms, Linear regression, Logistic regression, Naive Bayes and Classification trees. In this thesis, the students were divided into different performance types which was the basis for studying each performance trend on the algorithms. The evaluation showed that supervised machine learning can be used as a tool to help predict student performance. However the result obtained is dependent on knowing the performance types of students.