Relative movement of skin markers to underlying bone limits a valid interpretation of axial femorotibial rotation in noninvasive optoelectronic gait analysis. A distal femoral clamp is a practical solution for thigh marker placement, however, existing devices are still susceptible to measurement errors at increased angles of knee flexion. We developed the Femoral Epicondylar Frame (FEF), which should result in less femoral rotational measurement error due to its anatomic fitting and controlled pressure adjustment. Seven subjects with a total knee replacement in situ, mean age 71 years, mean body mass index 28, were equipped with the frame mounted with a set of tantalum markers. Fluoroscopic data was collected during a step-up motion. A three-dimensional model fitting technique was used to compare the in vivo position and orientation of the frame and the femoral prosthesis component of the prosthesis. The frame rotational measurement error appeared to be linearly dependent on the knee flexion angle. When considering knee flexion angles lower than 40° of flexion, the highest measurement error was 3.3° on average, with an absolute extreme of 6.2° It is concluded that the accuracy of the FEF is sufficient to evaluate axial knee rotation with optoelectronic gait analysis at group level in clinical studies. © 2010 Elsevier B.V.