Background and purpose Problems during knee extension, due to kinematic alterations, are not uncommon after total knee arthroplasty. Hinged prostheses provide higher stability than non-hinged designs and may minimize these alterations. Thus, in this in vitro study we investigated the quadriceps force required to extend the knee during an isokinetic extension cycle generating a constant extension moment after non-hinged and hinged total knee arthroplasty.Methods Human knee specimens were tested in a kinematic knee simulator under physiological conditions, after implantation of two types of non-hinged cruciate retaining prosthesis (Gemini; Link, Germany and Interax I.S.A.; Stryker, Ireland) and a hinged prosthesis (Rotations-Knie; Link, Germany). During simulation of an extension cycle from 120° knee flexion to full extension, the change in quadriceps force to produce the constant extension moment of 31 Nm was dynamically measured using a load cell attached to the quadriceps tendon.Results After implantation of the non-hinged prostheses, there was no alteration in maximum quadriceps force in knee flexion compared to physiological conditions, but alteration occurred at lower flexion angle (p = 0.002) and increased up to 1,257 (SD 273) N (p = 0.04) in knee extension. Following implantation of the hinged prosthesis, there was no alteration in quadriceps extension force in flexion but it decreased to 690 (SD 81) N (p = 0.003) in extension.Interpretation Hinged knee prostheses restore the quadriceps lever arm in knee flexion and improve the