The correlation between the insertion torque of a lumbar pedicle screw and the mechanical stability of the screw in the bone has been mentioned in in vitro studies. The purpose of this study was to confirm the factors affecting the insertion torque of such screws in vivo. Also, the contribution of insertion torque to the initial stability of the fusion area was to be analyzed in vivo. A series of 23 cases representing 50 lumbar vertebrae were included in this study, in which we examined bone mineral density using quantitative computed tomography (CT) prior to operation. Two screw shapes were utilized, with the insertion torque for each screw measured at two points in time. The correlation between insertion torque and mineral density was investigated. Screw positions were confirmed on postoperative CT scans, and the effect of the screw thread cutting into the cortex bone was investigated. Radiographic changes at three points during a period of 3 months were also measured, and we then evaluated the interrelations between these changes and insertion torque. Furthermore, the relation between insertion torque and instability at 3 months was investigated. Correlations of insertion torque and bone mineral density depended on screw shape. There was no correlation found with mineral density in the case of cylindrical screws. Insertion torque was not affected by the screw thread cutting into the cortex of bone. As for postoperative alignment changes, no definitive trends could be ascertained, and no interrelations with torque and alignment changes were observed. There is a possibility that insertion torque was related to early-stage stability, but no statistical relation could be determined.