This study presents a microanalysis of what information performers “give” and “give off” to each other via their bodies during a contemporary dance improvisation. We compare what expert performers and non-performers (sufficiently trained to successfully perform) do with their bodies during a silent, multiparty improvisation exercise, in order to identify any differences and to provide insight into nonverbal communication in a less conventional setting. The coordinated collaboration of the participants (two groups of six) was examined in a frame-by-frame analysis focusing on all body movements, including gaze shifts as well as the formal and functional movement units produced in the head–face, upper-, and lower-body regions. The Methods section describes in detail the annotation process and inter-rater agreement. The results of this study indicate that expert performers during the improvisation are in “performance mode” and have embodied other social cognitive strategies and skills (e.g., endogenous orienting, gaze avoidance, greater motor control) that the non-performers do not have available. Expert performers avoid using intentional communication, relying on information to be inferentially communicated in order to coordinate collaboratively, with silence and stillness being construed as meaningful in that social practice and context. The information that expert performers produce is quantitatively less (i.e., producing fewer body movements) and qualitatively more inferential than intentional compared to a control group of non-performers, which affects the quality of the performance.