Total knee arthroplasty aims to mimic the natural knee kinematics by optimizing implant geometry, but it is not clear how loading relates to tibio-femoral anterior-posterior translation or internal-external pivoting. We hypothesised that the point of pivot in the transverse plane is governed by the location of the highest axial force. Tibio-femoral loading was measured using an instrumented tibial component in six total knee arthroplasty patients (aged 65-80y, 5-7y post-op) during 5-6 squat repetitions, while knee kinematics were captured using a mobile video-fluoroscope. In the range of congruent tibio-femoral contact the medial femoral condyle remained approximately static while the lateral condyle translated posteriorly by 4.1 mm (median). Beyond the congruent range, the medial and lateral condyle motions both abruptly changed to anterior sliding by 4.6 mm, and 2.6 mm respectively. On average, both the axial loading and pivot position were more medial near extension, and transferred to the lateral side in flexion. However, no consistent relationship between pivoting and load distribution was found across all patients throughout flexion, with R-2 values ranging from 0.00 to 0.65. Tibio-femoral kinematics is not related to the load distribution alone: medial loading of the knee does not necessarily imply a medial pivot location.