Abstract
In comparing the high- stakes ESP tests administered in recent years in Iran for Master Degree and PhD levels admission purpose, a vast area of uncertainty arises as to the nature of scoring system used at the large scale for these competitive exams. Where we had the modular PhD entrance exam with a prerequisite EGP Module followed by a Specific Purpose Module, the participants of our Master Degree counterpart test sat for a joint or blended EGP-ESP subtest which then was scored and reported in a composite percentile and interpreted along with other knowledge subtests against a common national norm. In this article, we attempted to address the issues associated with these models of scoring and the problems which are due from it for our accountability system.