Carpal kinematics during a wrist flexion/extension motion using high-speed videodata acquisition was investigated. A cadaver forearm was stabilized, allowing unconstrained excursion of the wrist for passive range of motion (ROM). The extensor and flexor pairs of the wrist were looped together and a 1-lb weight was attached to each pair, simulating synergistic muscle tension. Capitate/radius and third metacarpal/radius angles were calculated to determine which measurement would be best for determining global wrist angle. The average difference in capitate/radius and third metacarpal/radius angles at each respective flexion/extension wrist angle for all wrists was 1.1 degrees +/- 1.6 degrees (the maximum difference was 4 degrees). Hence, the capitate-third metacarpal joint can be considered rigid. Capitate/lunate motion as described by capitate-radius Euler angles ranged from -16.9 to 23.5 with total capitate/lunate motion of 40.5 (35%) in the 114 degrees total global wrist ROM measured. Radius/lunate motion as described by lunate-radius angle ranged from -8.2 to 48.4 with total radius/lunate motion of 56.5 (49%) in the 114 degrees total global wrist ROM measured. During global wrist motion, the radiolunate joint contributes more motion in flexion than the capitolunate joint and the capitolunate joint contributes more motion in extension than the radiolunate joint. The instantaneous screw axes (ISAs) were calculated for each third metacarpal position with respect to the radius. The average distance difference between ISAs for the 4 wrists tested was -1.23 +/- 14.97 pixels. The maximum distance was 56.51 pixels and the minimum was -24.09 pixels. This new combination of motion analysis and 3-dimensional reconstructions of computed tomography images affords a high-speed, dynamic analysis of kinematics. It shows that during wrist flexion/extension, normal carpal kinematics does not have an ISA fixed in or limited to the capitate. In addition, the ISA data provide evidence that translational motion is a real and measurable component of normal carpal motion. These findings alter the understanding of carpal kinematics obtained from the results of previous studies which suggested that the center of rotation was fixed in the capitate.