The purpose of this study is to validate a noninvasive rotational knee laxity measuring device called "Rotameter P2" with an approach based on Computed Tomography (CT). This CT-approach using X-rays is hence invasive and can be regarded as a precise reference method that may also be applied to similar devices. An error due to imperfect femur fixation was observed but can be neglected for small torques. The most significant estimation error is due to the unavoidable soft tissues rotation and hence flexibility in the measurement chain. The error increases with the applied torque. The assessment showed that the rotational knee angle measured with the Rotameter is still overestimated because of thigh and femur displacement, soft tissues deformation, and measurement artefacts adding up to a maximum of 285% error at +15 Nm for the Internal Rotation of female volunteers. This may be questioned if such noninvasive devices for measuring the Tibia-Femoral Rotation (TFR) can help diagnosing knee pathologies and investigate ligament reconstructive surgery.