Although group dynamic assessment (GDA) has been gaining attention over the recent decade, its applicability in online context has been left rather underexplored. Hence, the current study examined the effects of GDA on developing EFL learners' written grammatical accuracy in the online context of 'Telegram'. To this aim, 60 Iranian EFL students whose age ranged from 14 to 18 years old were assigned randomly into two groups, namely GDA (N=30) and Non-GDA (N = 30). Initially, both groups wrote an essay on the same topic taken from Preliminary English Test (PET) as the pre-test. Afterwards, both groups wrote on seven writing tasks and shared them in their groups on Telegram. The students in the GDA group were provided with graduated and contingent feedback following a concurrent interactive approach. That is, the teacher offered a gamut of feedback according to the responsiveness of the students to fix errors. On the other hand, the non-GDA group only received direct corrective feedback without being afforded the opportunity to interact over their errors. Finally, both groups revised their own last essays as the first posttest and also wrote an essay on a new task. Results of the study indicated that the GDA group significantly outperformed the Non-GDA group in terms of gains measured by both revision and the new task. EFL teachers and teacher educators can be the beneficiaries of the current study to enrich their online sessions with more GDA-based feedback.