This study endeavored to examine the effects of corpus-based instruction on EFL students‟ academic writing skills, critical thinking skills, engagement, and perception. To conduct the research, a non-equivalent group pretest-posttest quasi-experimental design was employed. The experimental group (N=38) was instructed through corpus-based instruction; whereas, the comparison group (N=39) was instructed through the conventional approach. For data collection purposes, academic writing tests, questionnaire, textual analysis, semi structured interview, self-report survey, and reflective journal were used. The data which were gathered through academic writing tests, self-report survey and questionnaire were analyzed through one-way MANOVA, independent t-test, one-sample t-test, and paired sample t-test, and the qualitative data which were gathered through reflective journal, textual analysis, and semi-structured interview were analyzed qualitatively. Hence, the data indicated that students who were instructed their academic writing skills through corpus-based instruction outperformed the comparison group in overall academic writing skills (Wilks‟ λ =.299, F (4, 76) = 42.19, η2 = .701, which P<0.05). Among the four academic writing scales, the data confirmed that the students who engaged in the experimental group improved lexical resources (η2 = .495) and task achievement (η2 = .472) better than the other academic writing scales. Likewise, the textual analyses results also showed that the experimental group constructed better academic writing texts than the comparison group. The results obtained from tests and the textual analyses also proved that genuine linguistic features which were incorporated during the intervention contributed to improve the students‟ critical thinking skills. Moreover, the results gained from the self-report survey confirmed the students‟ better engagement in academic writing classes. The reflective journal also revealed concurrent results. Moreover, the results obtained from the questionnaire also indicated that the majority (m=4.3278) of students had positive reactions supporting the effectiveness of corpus-based instruction in academic writing classes. The semi-structured interview also confirmed students had positive perceptions towards corpus-based instruction. Therefore, academic writing teachers should consider corpus-based instruction and design materials and activities that satisfy the academic writing needs of EFL students