There is a strong case to be made for using telecollaboration in foreign language education: learners are engaged in regular, semi-authentic interaction; can develop strategies for learner independence (O'Rourke, 2007); be exposed to opportunities for negotiation of meaning (Kötter, 2003) and become aware of cultural differences in communicative practices (Belz & Kinginger, 2003), whilst at the same time exchanging with 'real informants' of the target culture and its behaviour (O'Dowd, 2013). In telecollaboration, several conditions may lead us to believe that learners are participating in a symmetric learning situation. They share a 'symmetry of status' (Zourou, 2009) as both interlocutors have the same status as language learners. Learners also spend equal amounts of time using each of the languages and we can imagine that there is 'symmetry of knowledge' (Dillenbourg, 1999) with participants sharing similar language levels. This paper details the evolution of a telecollaborative project's learning design and focuses on its asymmetrical nature before describing the students' learning experience. The practical study reports on a telecollaborative project run between Université Blaise Pascal (UBP) and the London School of Economics (LSE). The project brings together, every autumn semester, 8-10 English-as-a- foreign-language (EFL) students and 8-10 students of French-as-a-foreign-language (FFL). Both groups are of B2 CEFR level (Council of Europe, 2003). The exchange is centred on online synchronous sessions and the analysis of parallel texts. The presentation will initially focus on the evolution of the project's learning design since 2010 and what the teachers have learnt in terms of running the project. Three topics will be covered. Firstly, the evolution of the platform chosen for the online synchronous interaction that started out with a synthetic world before moving to a video-conferencing environment and then to an audio-graphic conferencing environment. The reasons for this evolution, the affordances of each for our learning design and hindrances for student learning will be discussed. Secondly, we will describe the evolution of the synchronous sessions' content and examine reasons for initially providing students with materials before asking students to source materials themselves. Thirdly, the evolution in the grouping of participants will be considered. The paper will then detail the current learning design and concentrate on its asymmetrical characteristics. Due to different institutional set-ups, the project is integrated into a two-year University language module for the EFL students whilst the FFL students partake voluntarily on a yearly basis. Consequences include different learning objectives and the fact that the project must be assessed for the EFL students but not for the FFL students. The different University calendars also allow a ten-week period to be devoted to the project by the EFL students but five weeks for the FFL students. We will outline how the learning design tried to reconcile this asymmetry. Our final focus will be the students' experience of telecollaboration. We will report on students' reactions to the learning design's evolution and, using post-project questionnaire data, the students' thoughts and feedback on the learning experience, including attitudes towards the asymmetrical design. We hope to offer practical leads concerning the learning design of telecollaborative projects and strategies for reconciling asymmetrical set-ups.