The use of transpedicular screw fixation has been widely accepted for the treatment of degenerative and traumatic pathology of the lumbar spine. Complications of spinal instrumentation can be serious. Screw misplacement can result in unintended durotomy, nerve root and/or cauda equina injury. In comparison to fluoroscopy-assisted screw placement, computer-assisted image guidance has been shown to achieve overall higher rates of accuracy. The O-arm is able to obtain computed tomography (CT)-type images with multiplanar reconstruction. In this study we evaluated a cohort of patients who underwent posterior lumbar fusion with pedicle screws utilizing the O-arm imaging system. A retrospective review of 40 consecutive patients who underwent posterior lumbar fusion surgery with O-arm utilization, was performed. The study population included 14 males and 26 females. Age range was 39-85 years with an average of 63.8 years. Twenty one patients had degenerative lumbar stenosis (52.5%) and 19 had spondylolisthesis (47.5%). Intraoperative CT-images were obtained. The mean time for surgery and screw placement was assessed. A total of 252 pedicle screws were sited using O-arm navigation system, with a mean of 6.3 screws per patient (range 4-10). On the basis of intraoperative CT, 3 screws were redirected, representing a 98.81% accuracy rate. The mean duration of surgery was 157.2 (90-240) minutes and the mean time for screw placement was 7.13 (3.08-15) minutes per screw. Three patients (7.5%) developed superficial wound infections which were treated conservatively. No patients required a return to the operating room because of screw malposition. The use of intraoperative O-arm imaging system with computer-assisted navigation significantly increases the surgical accuracy and safety of pedicle screw placement in lumbar fusion surgery.