Spine discectomy and fusion is a widely used surgical procedure to correct irreversible degenerative diseases and injuries to the intervertebral disk. The surgical procedure involves the removal of the damage disk material, the decortication of the fusion site, and the placement of the bone graft. Fusion is believed to generate additional stresses in the neighboring disks, which can subsequently lead to new disk degeneration and re-operation. The autologous bone has proven to be the best material for the fusion. However, the autologous bone has three major disadvantages: the high rate of donor site morbidity, the limited and sometimes poor quality of the amounts of bone available, and the extra operative time needed for harvest. For these reasons this study is undertaken to estimate the optimum amount of bone graft needed for a discectomy and correlate it to the change in stress in adjacent levels. A detailed and validated 3D finite element model of the complete human cervical spine (C1-T1) was altered to simulate segmental full and partial discectomies. One full fusion (bone graft occupies about 90% of the vertebral body) and seven partial fusions (bone graft occupies about 10%, 20%, 30%, 40%, 50%, 65%, and 75% of the vertebral body) were simulated at each of the four mid- and lower single levels of the cervical spine and the relationship between the change in stresses in the adjacent levels and the bone graft size (area) was studied. The changes in stress were compared with the previously obtained results of the unfused models. The fused and unfused models were preloaded with a 73.6 N compressive force representing the weight of the head and with a 1.5 Nm physiological moment in flexion, extension, lateral bending, and axial rotation. More than 132 cases were analyzed. The results showed that the necessary amount of bone graft needed for discectomy depends on the cervical disk level to be fused and varies between 30% and 75% of the disk area. The results also suggested that there is a threshold size of the bone graft area, before and/or after which, the long-term effects of the change in stresses in adjacent disks are biomechanically consequential.