While artificial intelligence in language learning has gained more attention, the effect of ChatGPT-integrated learning on English grammar remains largely underexplored. To address this significant gap, this study aims (1) to investigate the effect of ChatGPT-integrated learning on English grammar achievement and (2) to explore learners’ experiences and perceptions after engaging with ChatGPT to learn English grammar concepts. Sixty learners were involved in this quasi-experimental study, split into experimental and control groups. During the 6-week language instruction, the experimental group received ChatGPT-integrated grammar learning while the control group received non-ChatGPT-integrated grammar learning. Data were collected from pre-test, post-test, and semi-structured interviews. The quantitative data, the pre-test and post-test, were analyzed using the Mann-Whitney U Test, while the qualitative data, the semi-structured interview, were examined through thematic analysis. The result revealed that ChatGPT-integrated learning enhanced students’ grammar achievement. Although both groups enhanced their grammar scores, students from the experimental group obtained significantly higher scores than those from the control group. Most students had favorable perspectives on using ChatGPT in their English grammar learning instructions. However, some were found to worry about the likelihood of dependency and declined critical thinking. Careful planning of its pedagogical compatibilities, communication of potential risks and ethical considerations, and guidelines are needed to navigate the complexities involved in adopting ChatGPT into ELT classrooms and to mitigate the potential drawbacks of ChatGPT within the education field. This study, therefore, calls for policies and guidelines to integrate ChatGPT into curriculums.