Language tandem interactions provide a unique collaborative learning environment, as each participant takes turns being the native and the non-native side of the dialogue (Brammerts & Calvert, 2003). Although corrective feedback (CF) has received considerable attention in SLA literature (Lyster & Ranta, 1997, Sheen & Ellis, 2011), relatively little is known about CF occurring in these non-institutional peer-to-peer native/non-native interactions. We hypothesize that participants will mobilize resources that they share despite their different mother tongues and L1 cultures, namely non-verbal ones (prosody, gestures). Based on the qualitative yet systematic analysis in ELAN of 4 video recordings of interactions between French and English native speakers from the SITAF corpus (Horgues & Scheuer, 2013), we analyze CF focus, CF type, and the multimodal resources used for CF. Our study shows that CF is a highly multimodal activity (more than 86% of the time), identifies the main non-verbal resources used for CF request, provision and uptake and analyses the participants’ consistent idiosyncratic multimodal CF strategies.