Although research articles (henceforth RA) have been widely used in undergraduate teaching as assigned readings or supplementary teaching materials, studies addressing students' challenges in RA reading are infrequent. We conducted a two-cycle action research to examine the effect of genre instruction on RA reading comprehension of Year 3 and 4 undergraduate students in economics. In Cycle 1, students (n = 34) were introduced to move structures of conventional sections, including Introduction, Method, Results, Discussion, and Conclusion; in Cycle 2, students (n = 34) were introduced to a discipline-specific move model of economics. Apart from the conventional sections, the model used in Cycle 2 also describes the rhetorical structure of discipline-specific sections, such as Theoretical Model, Econometric Model, Robustness, and Mechanisms. Summary writing and multiple-choice tests were used to measure students' improvements in reading comprehension. The findings show that Cycle 2 instruction significantly improved students' performance in summary writing (medium effect size) and multiple-choice tests (small effect size), whereas Cycle 1 instruction did not. Also, fewer students in Cycle 2 perceived cognitive overload challenges than their peers in Cycle 1. The findings can be useful for EAP practitioners teaching similar cohorts of students.