Aims: The purpose of this study was to determine the intra-subject repeatability of the motion pattern of the PE inlay in a mobile-bearing total knee replacement (TKR) with respect to the post-op time. Methods: 75 mobile-bearing TKRs in 73 patients were included in this prospective study. Sagittal radiographs at 0°, 30°, 60° and maximum flexion were taken 3, 12, and 24 months post-op. On each X-ray, the AP position and the rotation of the PE inlay with respect to the tibial baseplate were determined based on a 2D algorithm. The accuracy of the method was ± 0.2 mm for the AP position and ± 1.7° for the angle of rotation. To classify the repeatability, the mean AP and rotation motion with respect to the flexion angle for each patient was computed and the overall standard deviation (STD) of all measurements with respect to the mean curves was calculated. The repeatability was defined as ‘excellent’ if the STD in the AP direction was less than 0.5 mm and the STD in rotation was less than 2.5°. It was defined as ‘good’ if the STD in the AP direction was less than 1 mm and the STD in rotation was less than 5°. Results: 19 of 75 knees (25%) showed an excellent repeatability and 33 of 75 knees (44%) displayed a good repeatability of the PE motion. Motion patterns were more repeatable between the 12 and 24 month results than between the 3 and 12 month results. Conclusions: The majority of the mobile-bearing knees exhibited a repeatable, patient specific motion pattern of the PE inlay. The fact that the repeatability was higher between 12 and 24 months may be attributed to a more stable state after rehabilitation. Mobile-bearing knees support patient specific motion in contrary to constrained fixed bearing knees.