Finite Element Models can be a helpful tool to investigate new, motion preserving means of treatment of spinal deformities. However, an effective validation is even more challenging than for a healthy spine, as in-vitro experiments are very hard to perform due to the limited availability of human cadaver specimens. Measuring the stiffness of scoliotic motion segments intra-operatively offers the possibility to acquire data in an in-vivo situation and at the pathology of interest. We developed a measurement setup to determine the lateral bending stiffness of motion segments after the opening of the back and before the corrective surgery of Adolescent Idiopathic Scoliosis. The apparatus includes a distraction forceps modified to measure force and displacement and an optelectronic camera to track motion. The combination of geometric computer models (vertebral models derived from MRI or CT), motion tracking and forceps data enables the determination of moment and angular displacement measurements for each distraction. The feasibility of this method was validated in a cadaver study using lumbar ovine spines. Measurements were performed with the distraction forceps and compared to equivalent experiments on a spinal loading simulator, as we consider this to be the gold standard for the determination of spinal stiffness. The stiffness computed with the proposed concept was within a range of 15% compared to data obtained with the spinal loading simulator under applied loads of less than 3 Nm. First, intraoperative trials have been performed. The preliminary results confirm the observation that the coupling is such that the anterior portion of the vertebra always rotates into the convexity of the curve when exposed to a left lateral bending.