The rise in L2 Intercultural Virtual Exchange (IVE) in recent years encourages more in-service teachers to incorporate such projects into their classroom via the implementation of three common task types: information exchange, comparison and analysis and collaboration. The introduction of IVE into initial teacher-training programmes is also gaining traction wherein student-teachers (STs) can experience it themselves and learn how to integrate IVEs into their future practice. This calls for a deeper understanding of task engagement with STs to help them to reflect on the task types which are successful (or less so) in IVE task design for their future learners. Although studies have reported L2 learner engagement in IVE projects (Dooly & O’Dowd, 2018), Gijsen (2021) argues that engagement in online tasks should be measured by examining the “special effort” students bring to a task (p. 59). She outlines 12 task parameters within behavioural, cognitive and emotional-attitudinal engagement including how learners manage the task, their interest in the topic and content, whether they consider the task appropriate, and the technology being utilised. Crucially, her research provides a means of assessing L2 learner online engagement in asynchronous and synchronous tasks. Along these lines, Reeve et al. (2021) categorise students’ agentic engagement as “how students proactively and constructively attempt to influence their instruction” citing such examples as the personalisation of and suggestions for altering tasks (p. 1). One means of verifying task and agentic engagement with STs in initial teacher-training programmes is via written reflections: Reflection being a critical element of teacher development (Farrell, 2022). Yet, there is little research to understand the relationship between task type, agentic engagement and reflective practice in initial teacher-training programmes. Therefore, this study examines firstly, which task types are reported as engaging by STs and, secondly, the extent that reflection and agentic engagement can be evidenced by this group dependant on task type. A telecollaborative exchange with pre-service L2 teachers at a French and Dutch university was carried out in Winter 2022. Student-teachers (n=33) completed three different task types on two different online platforms including introducing themselves through language biographies on the asynchronous video forum Flip (task 1), information exchanges about their teaching experience (task 2) and designing an IVE task for their future classroom (task 3) on the videoconferencing platform Big Blue Button. Following each task, students were asked to write a task reflection including their level of satisfaction regarding the task, communication, collaboration, intercultural experience and digital tools utilised. This study focuses on three student-teacher pairs (n=6). Data instruments included transcripts of video interactions, post-task reflections and a five-point Likert scale post-task questionnaire related to all three tasks to ascertain levels of learner engagement. Video interactions were transcribed and coded in ELAN. Participants’ post-task reflections were analysed using Bocquillion, Derobertmasure and Demeuse (2017). The post-task questionnaire provided descriptive statistics. Preliminary results suggest a high level of behavioural effort was exhibited under the engagement markers of topic and content, wherein participants expressed enjoying the topic, and partner orientation, i.e., paying close attention to the conversation with their partner. Regarding attitudinal-emotional effort, participants specifically engaged with the task parameter of technology in which they felt comfortable in the online environment. In terms of agentic engagement, French participants personalised the task more than their Dutch counterparts and suggested alternations to the task indicating deep levels of engagement. However, in examining participants’ post-task reflections, French participants tended towards more descriptive reflection when compared with their Dutch counterparts. Finally most participants (80%) cited the analysis and comparison task as the most engaging of the three. These preliminary findings shed light on the types of task that teacher-trainers can implement in future IVEs to foster task engagement. Further, the results of the post-task reflections outline recommendations for engendering deeper levels of reflection by student-teachers in initial teacher training programmes.