<p>In the digital education sphere, Language Massive Open Online Courses (LMOOCs) have garnered increasing attention, drawing millions of learners globally, a trend markedly accelerated by the COVID-19 pandemic. Despite their burgeoning popularity, research on the effectiveness of LMOOCs remains scant. Moreover, the prevailing research methodology largely hinges on narrative reviews from LMOOC platforms, with comparatively little direct engagement with the learners themselves. A notable gap exists in empirical investigations, particularly those adopting experimental or quasi-experimental designs, aimed at evaluating LMOOCs’ efficacy in bolstering second language (L2) acquisition, specifically in terms of grammatical range and accuracy. Addressing these lacunae, the present study aimed to assess the impact of LMOOCs on enhancing L2 learners’ grammar competence and their evaluations of these online courses. Utilizing a mixed-methods research approach including a quasi-experimental design with semi-structured interviews, this study involved fifty-seven participants, assigned into an experimental group (EG, <i>N</i> = 30) and a comparison group (CG, <i>N</i> = 27). While the EG enrolled in grammar-focused courses on Coursera, the CG did not receive any explicit grammar instruction. Contrary to expectations, no significant disparities in grammatical accuracy and range between the two groups were detected over time. Furthermore, the interview responses illuminated a general dissatisfaction with LMOOCs, revealing learners’ reluctance to pursue further language studies via these courses. The findings prompt a critical re-evaluation of LMOOCs’ instructional design and assessments as well as offer potential avenues for future research.</p>