A prospective clinical trial was performed to study the accuracy of pedicle screw placement of consecutive computer-aided and conventional techniques. Concerning the clinical performance of the navigation system, the average time of matching has been 8.4 minutes per vertebrae. For evaluation of the results, only complete intraosseous placement of the pedicle screw has been defined as correct position. Any medial or lateral perforation of the cortical structure of the pedicle was recorded as malplacement. In the CT-controlled patients complete intrapedicular placement of the screw was obtained in 36 of 45 thoracic (80%) and in 22 of 27 navigated lumbar (81%) pedicles. In the conventional cohort group 27 of 34 (79%) thoracal and 43 of 52 (83%) lumbar screws were completely in. No radicular neurologic damage, caused by a malplaced transpedicular screw has been observed in both groups. In the presented study is shown, that the application of the computer-assisted freehand navigation can improve results concerning the precision of spinal screw placement. Although, the spinal navigator has to consider a learning curve for the clinical inauguration of the system and the qualification of the implant system for computer-assisted application.