This study utilizing a sequential explanatory mixed-methods design explored the impact of mobile-based dynamic assessment (MDA) on EFL learners' writing skills. Three intact classes (N = 30), including intermediate EFL learners attending a private English language school in Iran, participated in this study. They were evenly divided into two experimental and one control groups. The DIALNG online diagnostic test was used to assess the participants' written proficiency and also as an instrument to collect the pre-and post-test scores. During the treatment sessions, the students were required to complete writing tasks over their Google Docs mobile app shared with the instructor. She provided text- and voice-based mediations to the experimental groups following an interactionist DA using both WhatsApp and Google Docs. Follow-up interviews were conducted to assess the experimental groups' perceptions towards each mediation type. T-test and ANOVA along with the thematic analysis were used to analyze the quantitative and qualitative data, respectively. The findings showed that only the text group's post-test scores significantly improved and there was a significant difference among the three groups in their post-test scores which indicated outperformance of the voice group. The overall results showed that MDA enhances EFL learners' written proficiency as a result of the collaborations between the learners and the instructor using text and voice-based mediation. The thematic analysis of data revealed the participants' satisfaction with both mediation types in terms of being efficient, convenient, and causing less social pressure.