Collaborative remembering is a joint activity that involves the establishment and reinforcement of a common ground. In this study, we investigate some ways in which collaborative remembering involves interactive coordination of non-verbal behaviors. Our data consist of video recordings of small groups of people that are reconstructing holiday memories together. For each 500 ms, these videos have been annotated in terms of the participants' bodily behavior and posture, including variables such as manual gesture, shoulder shrugs, leaning direction and gaze. We compared instances of 'simultaneous alignment' (two or more people concurrently performing the same behavior) and 'sequential alignment' (two people performing the same behavior in short temporal succession) to chance baselines, and found that the latter is more common that the former. Our analysis furthermore suggests that the degree to which participants coordinate their behaviors is stable across the course of the conversation (i.e., time-independent), but connected with specific activities within the larger discourse of collaborative remembering.