The present study investigated how constraining movement affects interpersonal coordination and joint cognitive performance. Pairs of participants worked cooperatively to solve picture-puzzle tasks in which they conversed to identify differences between pictures in 3 degree-of-constraint conditions: both participants were free to move their hands (free-free; FF); both participants' hands were restrained (restrained-restrained; RR); and the hands of 1 participant were free while the hands of the other participant were restrained (free-restrained; FR). Eye tracking data were collected, and movement was measured at the waist, hand, and head. Data were analyzed using Cross-Recurrence Quantification Analysis (CRQ). Postural sway coordination, gaze coordination, and task performance were predicted to be highest in FF, followed by RR, and then by FR. Results showed the asymmetric FR condition generally exhibited lesser degrees of coordination than the symmetric Conditions FF and RR, and that the patterning of coordination in the symmetric conditions varied across the measured body segments. These results demonstrate that movement restraints affect not only interpersonal postural coordination, but also joint attention. Additionally, significant positive relationships were found between task performance and total amount of anterior-posterior movement measured at the head, hand and waist; number of utterances; and number of differences pairs found in the puzzles. These findings indicate a relationship between movement and task performance consistent with the hypotheses that both interpersonal coordination and cognitive performance are sensitive to local action constraints.