OBJECTIVE: To evaluate the feasibility and safety of three-dimensional (3D) printed drill guide template-assisted percutaneous pedicle screw fixation for multiple-level thoracolumbar fractures. METHODS: Clinical data of 19 patients with multilevel thoracolumbar fracture without nerve injury who underwent surgical treatment between May 2017 and January 2019 were retrospectively analyzed. There were 9 males and 10 females and their age ranged from 22 to 63 years, with an average age of 43.6 years. Injury cause included traffic accident injury in 12 cases, and fall from height injury in 7 cases. A total of 40 fractured vertebrae were involved in T to L levels. According to AO classification, there were 29 fractures of type A1, 9 fractures of type A2, and 2 fractures of type A3. According to TANG Sanyuan classification, multiple-segment thoracolumbar fractures were classified as 17 cases of type ⅠA, 1 case of type ⅠB, and 1 case of type ⅡC. The time from injury to operation was 2-6 days, with an average of 3.1 days. The 3D-printed universal drill guide template was used for assisting percutaneous pedicle screw fixation during operation. Intraoperative blood loss, average operation time and fluoroscopy frequency of each screw were recorded. Visual analogue scale (VAS) score was used to evaluate the improvement of low back pain before operation, at 3 days after operation, and at last follow-up. According to the CT at 3 days after operation, the Gertzbein and Robbins scales were used to evaluate the accuracy of screw insertion (the grade A and grade B were regarded as accuracy, the grade A was regarded as excellent of screw insertion). The Cobb angle in sagittal plane of the fracture segment was measured, and the percentage of anterior edge of injured vertebral height was calculated. The consistency of the inclination of bilateral pedicle screws were analyzed postoperatively, and compared the angle of the intraoperative guide plate with the inclination of screw to verify the effectiveness of the guide plate in controlling the inclination. RESULTS: All the 19 patients completed the operation successfully, and the intraoperative blood loss was 44-67 mL, with an average of 54.3 mL. The average operation time for each screw insertion was 7.3-11.1 minutes, with an average of 9.6 minutes. The average fluoroscopy frequency of each screw insertion was 1.6-2.5 times, with an average of 2.0 times. No spinal cord, nerve root injury, infection, and other complications occurred. All patients were followed up 24-38 months, with an average of 28.7 months. The accuracy of pedicle screws was evaluated by using Gertzbein and Robbins scales: 145 screws were grade A and 11 screws were grade B. The accuracy of screw insertion was 100% and the excellent rate was 92.9%. The CT data at 3 days after operation showed no significant difference in the inclination between the left and right screws in the same vertebral body ( =0.93, =0.36). There was no significant difference between the angle of guide plate and the screw inclination ( >0.05). The VAS score, Cobb angle in sagittal plane, and the percentage of anterior edge of injured vertebral height were significantly improved at 3 days after operation and at last follow-up, and the VAS score was declined at last follow-up compared with 3 days after operation, all showing significant differences ( <0.05). There was no significant difference in the sagittal Cobb angle and the percentage of anterior edge of injured vertebral height between two postoperative time points ( >0.05). At last follow-up, no internal fixators were loosened or broken, and all fractures healed well. CONCLUSION: For the multiple-level thoracolumbar fractures, 3D-printed drill guide template assisting percutaneous pedicle screw fixation can reduce the operation time, intraoperative blood loss, and fluoroscopy frequency and the screw insertion is accurate and has a good reduction effect.