Objective To investigate the effectiveness of spinal canal decompression with microendoscopic disectomy (MED) and pillar vertebral space insertion through pedicle of vertebral arch for thoracolumbar neglected fracture. Methods Between February 2006 and November 2009, 30 patients with thoracolumbar neglected fracture were treated by spinal canal decompression with MED and pillar vertebral space insertion through pedicle of vertebral arch. There were 22 males and 8 females with an average age of 36.2 years (range, 17-58 years). The disease duration was 6 weeks to 14 months with an average of 5.3 months. All patients had single vertebral compression fracture, including T9 in 1 case, T11 in 2 cases, T12 in 5 cases, L1 in 11 cases, L2 in 5 cases, L3 in 5 cases, and L4 in 1 case. The preoperative Cobb angle was (27.5 ± 7.5) ° . The preoperative height of vertebrae was (26.67 ± 5.34) mm. The visual analogue score (VAS) was 5.8 ± 1.4. According to Wolter classification for spinal canal stenosis, there were 17 cases of grade 1, 10 cases of grade 2, and 3 cases of grade 3. According to Frankel grade, 3 cases were in grade A, 8 cases in grade B, 13 cases in grade C, and 6 cases in grade D. Results The average operation time was 70 minutes (range, 40-120 minutes) and the average blood loss was 180 mL (range, 100-400 mL). The hematoma occurred in 1 case, and other incisions healed by first intension. No deep vein thrombosis of the lower extremity occurred. All patients were followed up 26 months on average (range, 24-46 months). The Cobb angle and vertebral height at 3 days and last follow-up were significantly improved when compared with ones before operation (P lt; 0.01). At last follow-up, the spinal canal stenosis was grade 0 in 27 cases and grade 1 in 3 cases according to Wolter classification. At 24 months after operation, the spinal function was obviously improved; 1 case was in grade A, 1 case in grade B, 3 cases in grade C, 9 cases in grade D, and 16 cases in grade E according to Frankle grade, showing significant differences when compared with preoperative ones (P lt; 0.05). The VAS score at 1 month after operation was significantly higher than that before operation (P lt; 0.01), then the score showed downtrend along with time, and it was significantly lower at 24 months after operation than before operation (P lt; 0.01). Conclusion Spinal canal decompression with MED and pillar vertebral space insertion for thoracolumbar neglected fracture has short surgical time, less blood loss, and satisfactory reduction, but higher technical requirement is necessary for MED.