Intraoperative spinal sonography to check whether fragments have been successfully repositioned requires enlargement of the natural interlaminar window to 8 x 10 mm. The present study was performed to measure any rotational instability caused by such a laminotomy and compare it with that resulting when the conventional methods of checking the spinal canal, e.g. hemilaminectomy and laminectomy, are applied. The investigations were carried out in 10 human vertebral columns, in the area of T-12, L-2 and L-4. Torsional loads of up to 20 Nm were applied to the spines and axial rotations were recorded. Laminotomy, hemilaminectomy and laminectomy were done step by step. The loading-unloading cycles were achieved with initial axial rotation and decreasing torsional moment after measurement the dorsal manipulations. It was shown that interlaminar fenestration causes a hardly measurable instability. A laminotomy to 10 x 20 mm, allowing direct repositioning of the posterior vertebral surface, causes a 6% loss of stability. A significant decrease is caused by hemilaminectomy, with a 20% loss and by laminectomy, with 27% loss of rotational stability compared with the intact spine. Rotational instability of the spine causes severe pain and often has to be treated by spondylodesis. The present investigations show that the spinal canal should not be checked by hemilaminectomy or laminectomy. For this we now use intraoperative ultrasound.