Laminectomy and discectomy for herniated nucleus pulposus has been a common operative procedure, but about 15-40% of patients undergoing these very common procedures have unsatifactory results. Three are many causes of failed back syndrome and postoperative peridural fibrosis accounts for about 10-1596 of failed back syndrome. There have been many reports about the formation and prevention of peridural fibrosis and two theories have been proposed as to the origin of peridural or epidural fibrosis. However, there was one big deficiency in that none of the studies on animals truly replicated the surgical procedures performed on humans. The purpose of this study is to investigate the sourece and evolving pattern of postlaminectomy scar, and to provide basis for studies that will be done to prevent scar adhesion of nerve roots after laminectomy and discectomy. Sixty rabbits weighing between 2.5 and 3.0Kg were used. A routine laminectomy of 4th and 5th lumbar vertebrae was performed and the ligamentum flavum and epidural fat tissue were excised carefully. After discectomy, primary closure of wound was done in 30 rabbits as an control group. A primary closure was done after covering the laminectomy defect with Dacron sheet in 30 rabbits of the experimental gorup. The animals in the two groups were sacrificed at varying intervals of three days, one, three, six, and twelve weeks after the operation. Whole mount histologic sections at the disc between 4th and 5th lumbar vertebrae were prepared. The histologic findings lead us to summarize that; 1. Hemorrhage occurred markedly on the posterior side of dura in the control group but Dacron sheet prohibited hematoma to extend into the spinal canal in the experimental gorup. Hemorrhage was absorbed in all cases within three weeks after the operation. 2. Invasion of collagen fiber occurred from one week after the operation. Scar adhesion to the dura was most prominent of the posterior aspect of the dura in the control group, however, in the experimental gorup scar adhesion to the dura was prevented by Dacron sheet. 3. On the anterior side of the dura, scar tissue orginiated from the anulus fibrosus adhered to the dura and nerve root in both groups. The grade of adhesion increased with it was most dense at twelve weeks' duration. 4. Regarding to the extension of scar into the spinal canal and nerve root adhesion, there was no definite statistical difference between two groups. In conclusion, the scar tissue around the dura and nerve root was originated from the anulus fibrosus and laminectomy site. Dacron sheet may act as a barrier to prevent the scar extension form the laminectomy site but not from the anulus fibrosus, and more experimental and clinical studies should be done to prevent scar adhesion to the nerve root anteriorly.