Criterion referencing, which has been fully launched in the University, evaluates student performance in an assessed work against a predetermined set of goals and standards. Scoring student performance seems unnecessary. Yet, when multiple assessments are involved in a course, the grades of individual assessments are commonly converted into numeric scores whose weighted sum are then converted back to a course grade. A grade-score conversion is thus called for. This presentation addresses two technical issues when combining student performance in all assessments into a course grade. Firstly, the choice of grade-score conversion table is non-trivial especially in a course with multiple sections. Secondly, from the mathematical point of view, some choices of predetermined cut-off scores are preferable so as to reduce the chance of rampant grade inflation or deflation.