Adequate fixation is fundamental in revision total knee arthroplasty; consequently, surgeons must determine the correct set-up for each patient, choosing from numerous stem solutions. Several designs are currently available on the market, but there are no evidence-based quantitative biomechanical guideline yet. Therefore, several stems were designed and analyzed using a previously-validated finite-element model. The following parameters were studied: stem design characteristics (length and shape), added features (straight/bowed stem), fixation technique, and effect of slots/flutes. Bone stress and Risk of Fracture (RF) were analyzed in different regions of interest during a squat (up to 120°). For the femoral stem, the results indicated that all parameters influenced the bone stress distribution. The maximum von Mises stress and RF were always located near the tip of the stem. The long stems generated stress-shielding in the distal bone. Regarding the tibial stem, cemented stems showed lower micromotions at the bone-tibial tray interface and at the stem tip compared to press-fit stems, reducing the risk of implant loosening. The results demonstrated that anatomical shapes and slots reduce bone stress and risk of fracture, whereas flutes have the opposite effect; no relevant differences were found in this regard when alternating cemented and press-fit stem configurations. Cemented tibial stems reduce antero-posterior micromotions, preventing implant loosening.