PURPOSE: The aim of this study was to quantify the contributions of medial soft tissues to stability following cruciate-retaining (CR) or posterior-stabilised (PS) total knee arthroplasty (TKA). METHODS: Using a robotic system, eight cadaveric knees were subjected to ±90-N anterior-posterior force, ±5-Nm internal-external and ±8-Nm varus-valgus torques at various flexion angles. The knees were tested intact and then with CR and PS implants, and successive cuts of the deep and superficial medial collateral ligaments (dMCL, sMCL) and posteromedial capsule (PMC) quantified the percentage contributions of each structure to restraining the applied loads. RESULTS: In implanted knees, the sMCL restrained valgus rotation (62 % across flexion angles), anterior-posterior drawer (24 and 10 %, respectively) and internal-external rotation (22 and 37 %). Changing from CR TKA to PS TKA increased the load on the sMCL when resisting valgus loads. The dMCL restrained 11 % of external and 13 % of valgus rotations, and the PMC was significant at low flexion angles. CONCLUSIONS: This work has shown that medial release in the varus knee should be minimised, as it may inadvertently result in a combined laxity pattern. There is increasing interest in preserving constitutional varus in TKA, and this work argues for preservation of the sMCL to afford the surgeon consistent restraint and maintain a balanced knee for the patient.