OBJECTIVE: The direct repair of the pars interarticularis defect using a lag screw with bone grafting(Buck operation) is an anatomical, less invasive surgical treatment preserving motion segment. This paper reports the methods and long-term results of the direct screw repair of the pars interarticularis defect and fusion in symptomatic lumbar spondylolysis using a cannulated lag screw. METHODS: Ten patients with symptomatic spondylolysis nonresponsive to medical treatment more than 6 months underwent this operation. Eight patients(6 men and 2 women) were followed for longer than 24 months after surgery. Mean age at the time of surgery was 28.3(18-43) years. The involved lumbar vertebra were L5 in 6 cases, L3 in 1 case and L4 in 1 case. All patients except 1 case had bilateral lesion. All patients underwent preoperative magnetic resonance(MR) image and plain radiographs including functional view. Patients with slippage, instability, moderate to severe intervertebral disc degeneration on preoperative MR image at the affected segment and isthmus defect more than 4mm in length were contraindications to this procedure. Surgical outcome was evaluated with visual analogue scale(VAS), Macnab classification, and functional rating index(FRI). Postoperative plain radiographs were taken 4 weeks, 3 months. 6 months, 12 months, 18 months, 24 months after surgery and postoperative computed tomography scan was performed between 12-24 months after surgery. RESULTS: The average length of inserted screws were 42mm(35-45mm). Three cases experienced Excellent outcome, Good in 5 cases according to Macnab classification 6months after operation. Preoperative mean FRI scrore is 86.0%(34.4points) and postoperative 24months is 32.8%(13.1points). DeltaFRI is 0.62. There was no operation related complication including implant related complication(eg. screw fracture, screw malposition). One case with unilateral pars defect showed fusion as early as 6 months. CONCLUSION: These results indicate that Buck operation is a reliable and safe operation for symptomatic lumbar spondylolysis. Long-term data are needed for further evaluation for fusion.