Anatomical measurements were studied on 40 dry axis vertebrae to determine the suitability of the groove for the vertebral artery for atlanto-axial transarticular screw fixation technique. We measured 13 parameters including three angular and 10 linear dimensions related to the groove of the vertebral artery, pedicle, and pars interarticularis and evaluated 80 measurements for each parameter. All measurements were done after placing a Kischner guide win through the pedicle. We found that differences between measurements on the left and right sides of each vertebra were nonsignificant. In spite of the variability in measurements such as height, width, and median angle of the pedicle, the decline angle fur instrumentation, the depth of the groove fur the vertebral artery, and the internal height of the pars interarticularis, all of these had good symmetry. However, there were statistically significant differences between the sides in measurements for both the width (P=0.05) and the angle (P<0.02) of the pedicle allowing instrumentation and they did not show good symmetry. The risk of vertebral artery injury was found to be 22.5% per specimen, or 16.25% per screw inserted because the internal height of the pars interarticularis at point of fixation was less than or equal to 2.1 mm. In addition. We found that the pedicle width allowing instrumentation was not suitable in 12.5% of screws inserted because their values were less than or equal to 6 mm. When the width of the pedicle for instrumentation and the internal height of the pars interarticularis were both evaluated together, we also found that this technique would he extremely dangerous in 7.5% of specimens. In conclusion, the internal height of the pars interarticularis and the width of the pedicle for instrumentation should be evaluated together in thin CT sections preoperatively, because of the risk of vertebral artery injury in patients upon which atlanto-axial transarticular screw fixation is to be performed. (C) 2000 Wiley-Liss, Inc.