Introduction. Beneath infection, instability and malalignment, aseptic tibial component loosening remains a major cause of failure in total knee arthroplasty (TKA) [1]. This emphasizes the need for stable primary and long-term secondary fixation of tibial baseplates. To evaluate the primary stability of cemented tibial baseplates, different pre-clinical test methods have been undergone: finite element analysis [2], static push-out [3,4] or dynamic compression-shear loading [5] until interface failure. However, these test conditions do not reflect the long-term endurance under in vivo loading modes, where the tibial baseplate is predominantly subjected to compression and shear forces in a cyclic profile [5,6]. To distinguish between design parameters the aim of our study was to develop suitable pre-clinical test methods to evaluate the endurance of the implant-cement-bone interface fixation for tibial baseplates under severe anterior (method I) and internal-external torsional (method II) shear test conditions. Materials & Methods. To create a clinically relevant cement penetration pattern a 4. th. generation composite bone model was customised with a cancellous core (12.5 PCF cellular rigid PU foam) to enable for high cycle endurance testing. VEGA System. ®. PS & Columbus. ®. CRA/PSA ZrN-multilayer coated tibial baseplates (2×12) were implanted in the customised bone model using Palacos. ®. R HV bone cement (Figure 1). An anterior compression-shear test (method II) was conducted at 2500 N for 10 million cycles and continued at 3000 N & 3500 N for each 1 million cycles (total: 12 million cycles) simulating post-cam engagement at 45° flexion. An internal-external torsional shear test (method II) was executed in an exaggeration of clinically relevant rotations [7,8] with ±17.2° for 1 million cycles at 3000 N tibio-femoral load in extension. After endurance testing either under anterior shear or internal-external torsion each tibial baseplate was mounted into a testing frame and maximum push-out strength was determined [3]. Results. The cement penetration depth and characteristic pattern were comparable to 3D-CT scans of 24 cemented human tibiae from a previous study [5]. From the final push-out testing, no statistical significant differences could be found for anterior compression-shear testing (method I) with VEGA System. ®. PS (2674 ± 754 N) and Columbus. ®. CRA/PSA (2177 ± 429 N) (p = 0.191), as well as internal-external torsional shear testing (method II) between VEGA System. ®. PS (2561 ± 519 N) and Columbus. ®. CRA/PSA (2825 ± 515 N) tibial baseplates (p = 0.399). Discussion. The newly developed methods allow the evaluation of the endurance behaviour of the implant-cement-bone interface fixation for tibial baseplates in comparison to clinically long-term established knee systems, based on a combination of a suitable artificial bone model and severe anterior and internal-external torsional high cycle shear test conditions