There were many methods of assessing the time for fusion after fusion operation:clinical and radiological ones. Recently CT and MRI are also utilized for the more accurate assessment of fusion time together with serial simple X-rays and the tomograms. However, it is difficult to define the exact time for union by the simple X-ray alone with certain intervals, if the X-rays are not taken every week unitl the diagnosis of clinical union can be clearly made. For this study after Robins-on and Smith or Baily and Badgley's procedures, simple X-rays which were taken with certain inteval were used. Material and method: Sixty-eight patients underwent an anterior lower cervical interbody fusion by untilizing the iliac graft from January 1980 to December 1989. Among these 68 patients, only 28 could be evaluated with sufficient and roentgenographical follow-up, excluding the patients having an additional plate fixation from this sutdy. One segment fusion in 13 cases, two-segment fusion in 14 cases, and threesegment fusion in one case were done. Some graft extrusion (less then 20% of its width) was allowed to observe the absorption of the extruded portion. As for impacting the iliac graft, the round cortical edge of the iliac cortex was placed posteriorly and the cancellous edge anteriorly, because the extruded cancellous edge was much easily absorbed. When cervical instability was present, bed rest under cervical traction for 6 to 8 weeks after surgery, was prescribed which was followed by halo-vest, and when the instability was absent, patients were mobilized earlier. In each patient, disc height, joint angle, adaptive changes of bone graft and development of corporal inwaisting of the fused segment were observed by postoperative roentgenograms to assess the union. Results:1. The average loss of disc height was 4.8mm during the period of observation. The average changes of the joint angle, measured by Cobb's method, was 5.1 degrees. 2. The inwaisting sign apperared from postoperative 9.9 weeks on an average (6 to 28 weeks). The time of detection of this inwaisting sign coincided with the time of stabilization of disc height and joint angle. 3. In cases of two motion segment fusion with a piece of large graft had less graft collapse than that of individual motion segment fusion with two independent grafts, and the corporal inwaisting sign observed earlier in the former procedure. Gradual extrusion and/or crumbling of a graft mostly ended up with delayed or non-union, while gradual absorption of the extruded portion of the graft without further extrusion or crumbling ended up with successful fusion. Discussion and Summary:Upto now, in assessing or evaluating the time for fusion after anterior cervical interbody. fusion with bone graft, simple X-rays have been used in most instances with certain accuracy. Therefore it can be said that the time for fusion was inaccurately assessed with a wide range of errors. To make more accurate diagnosis of corporal fusion, serial-X rays were taken with two week interval from 4 weeks after surgery upto 20 weeks. In most instances in the past, time for fusion was assessed only by the figure of the graft, disc space change and collapse of the vertebral body, and attention was not paid much to the absorption of the extruded portion of the un-crumbled graft and the change of the joint angle. Gradual absorption of the extruded portion of the graft takes place early by the process of remodelling, the "so-called inwaisting phenomenon', when the graft is taken to its graft bed. It is suggested that the more accurate time for fusion after anterior fusion can be made by observing the radiological changes of the disc space and the joint angle, the collapse of the vertebral body, and the process of the absorption of the extruded portion of the graft. Conclusion:The maintenance of the disc height and joint angle without further change, and inwaisting sign of the fused segment are the most important factors favoring the prognosis of the union of the fused segment.