The purpose of our research is to develop surgical navigation systems to enhance surgical safety. Our systems make use of augmented reality technology to superimpose, on the surgery screen on a real time basis, patients' organ models reconstructed in 3D from their X-ray CT data taken before surgery. By doing so, the systems display anatomical risk materials, tumors and blood vessels which surgeons cannot see with their naked eyes. This will in turn lead to surgeons intuitively grasping the inner structures of the operational fields. We so far have been developing navigation systems that can conduct surgeries in various fields. The basic structure of the navigation systems are the same. The navigation systems uses different peripheral equipment and different methods to display navigation images which best meet the demands of each type of surgery. In this thesis, we report on our navigation systems for 2 types of surgery - endoscopic sinus surgery and hepatobilialy-pancreatic surgery.