Background Minimally invasive techniques are gaining wide-spread application in lumbar fusion surgery, because they may have advantage over conventional open surgery in approach-related morbidity. This research was aimed to evaluate the safety and accuracy of the techniques of minimally invasive transforaminal lumbar interbody fusion by using a computer-assisted spinal navigation system combined with electromyography monitoring. Methods Sixteen patients underwent minimally invasive transforaminal lumbar interbody fusion. A computer-assisted spinal navigation system and electromyography were used for guiding pedicle screw placement. The operative duration, blood loss, complications, and fluoroscopic time were recorded. Clinical outcome was assessed by Visual Analog Scale and Oswestry Disability Index. Radiographic images were obtained to evaluate the accuracy of pedicle screw placement and fusion rates. Results The Visual Analog Scale and Oswestry Disability Index scores were vastly improved postoperatively. A total of 64 pedicle screws were implanted and three were regarded as misplacement by post-operative CT scan. Three screw trajectories were adjusted according to intra-operative stimulus-evoked electromyography monitoring. The average fluoroscopy time in each patient was 31.8 seconds, which equals to 7.9 seconds per pedicle screw. No patients had instrument related neurological complications, infection, implant failure or revision. Successful fusion was found in all patients. Conclusions The combination of navigation system and real-time electromyography monitoring can make the minimally invasive operation more safe and accurate while decreasing radiation exposure time of the medical staff and patient and minimizing the chance and the degree of the pedicle screw misplacement. Chin Med J 2012;125(22):3947-3951