The article presents the results of the experimental enquiry aimed at studying the impact of cooperative learning on improving grammar skills. The research focused on two interaction modes -- pair work and group work, which were compared in terms of their potential to address accuracy. The experiment was carried out in the first term of the 2020-2021 academic year (September-October). The participants included 23 third-year students of the Foreign Philology Faculty of H. S. Skovoroda Kharkiv National Pedagogical University and constituted two academic groups, with one being offered pair-work activities and the other working in small groups. The experimental study comprised three phases: pre-experiment testing, teaching, post-experiment testing. Both quantitative and qualitative methods were used. The results demonstrate the validity of the assumption about the efficacy of pair work and group work in promoting accuracy, at the same time indicating no statistically significant difference between the two modes. The presented evidence suggests that grammar instruction can clearly benefit from incorporating both pair work and group work into classroom practices, however further research might be advisable to maximize the effectiveness of these interaction patterns by refining the procedure involved and improving students' cooperative learning skills.