Many earlier studies assessed virtual vs. traditional learning outcomes. The majority of them had two limitations. (1) learners self-selection of virtual classes, and (2) the lack of exams proctoring. It has been stated that these factors give more opportunities of unrealistic elevation of learning outcomes of virtual classes over the traditional ones. This study is of comparative corpus-based nature applied on 1324 male students of Medicine enrolled in Joint First Year Program (JFYP) at Umm Al-Qura University (UQU) in the first semesters of the academic years (2017-2018) and (2018-2019). Participants of this study were given English classification test before commencing their JFYP. In the first semester of the academic year (2017-2018), 624 students were taught English in traditional classes by the English Language Center (ELC) at UQU. The other 700 students studied English through virtual classes by TeachCast with Oxford via Eleutian platform. Then, all participants sat for final exams by the end of their first semesters. Comparing the participants’ grades in these two tests is meant to (1) test the claims of the earlier studies, (2) reveal the impact of English instruction by the ELC at UQU and (3) disclose whether the English learning outcomes of controlled virtual classes surpass the traditional classes’ or not. Findings of this study indicated that the learning outcomes of virtual classes significantly surpassed the traditional classes although the affecting factors stated by earlier studies were eliminated. Then, the study recommendations were suggested.