This mixed-methods study examined the efficiency of a student-selection peer review assignment protocol as compared to the typically implemented teacher-assigned-pair protocol while using L1 as a mediational tool. The participants were 60 undergraduate medical students attending a general English language proficiency course. The students were assigned into two experimental groups of teacher-selected (the teacher assigned students into pairs) and student-selected (the students were free to select peer work) and a control group (there was no pair and the students work individually). Working on some lexical language-related episodes (LREs), both teacher-selected and student-selected groups were allowed to use L1 freely while using L1 was forbidden for the control group. Using a sequential mixed-method design, the quantitative results of the ANOVA indicated that teacher-assigned and student-selected pairs using L1 demonstrated better domain learning outcomes over lexical LREs and outperformed the control group. In order to analyze the qualitative data, thematic analysis was used. The thematic analysis of learners’ LREs and portfolio revealed two themes in relation to the learners’ use of L1 and pair-formation method: (1) effectiveness of L1 use over English use, and (2) effectiveness of pair-formation over individual learning of English. The results indicated that L1 use in pairs could enhance collaborative chances during tasks in language classes. Finally, pedagogical implications are offered.