The present study made an effort to determine the impacts of proactive and retroactive meta-cognitive strategy instruction on Iranian ESP learners’ autonomy. Furthermore, it strived to examine the degree to which the interaction between this instruction and proactive/retroactive instruction conditions influenced these learners’ autonomy in their courses. To this end, first, the researchers selected 168 intermediate-level ESP learners from among the ESP learners of Islamic Azad University (Tabriz Branch) as the participants. Second, they non-randomly assigned the participants to four groups including the proactive experimental group, retroactive experimental group, proactive control group, and retroactive control group. Third, the researchers administered the autonomy pretest to all of the groups. Fourth, they used the Adobe Connect Learning Management System to provide the proactive experimental group, and the retroactive experimental group with their relevant treatments in ten sessions. Nonetheless, they used traditional language instruction techniques to provide the control groups with their instruction. Fifth, the researchers administered the autonomy posttest of the study to the participants after the end of the treatment sessions. Finally, they used SPSS 24 to analyze the data. The results showed that meta-cognitive strategy instruction ameliorated the participants’ autonomy. Moreover, the retroactive instruction condition was more efficacious than the proactive condition. In addition, the interaction between meta-cognitive strategy instruction and retroactive condition had a significant positive effect on the learners’ autonomy development. The results may provide ESP teacher educators, syllabus designers, and instructors with guiding principles regarding meta-cognitive strategy instruction. تفاصيل المقالة