Objective To explore the efficacy of early aggressive debridement with implant retention, primary wound closure, closed suction drain without irrigation and antibiotic therapy for the treatment of delayed deep infection after spinal fixation. Methods 4 057 patients were underwent dorsal spinal fixation from January 2010 to June 2014. Among them, 42 cases of delayed deep infection after operation were included in the study. There were 25 males and 17 females, with an average age of 68.6±8.1 years (ranged from 53 to 83 years). The diagnosis of delayed deep infection was based on the time of onset, clinical symptoms and signs, imaging and laboratory findings. Surgical debridement was performed immediately after diagnosis of infection. In addition, devitalized and necrotic tissue and biofilms which adhered to the surface of the implant were removed meticulously and thoroughly. Primary wound closure was performed in each patient, and closed suction drains were maintained for about 7-10 d without irrigation. Routine sensitive antimicrobial drugs was applied for 3 months after operation. Results 42 cases were all followed up for 24 to 72 months with an average of 46 months. Among the 42 infected patients, 3 patients were underwent posterior cervical spine surgery and 39 patients were underwent posterior lumbar spine surgery. There were 13 cases of staphylococcus aureus infection, 7 cases of escherichia coli infection, 3 cases of ESBL escherichia coli infection, 3 cases of enterobacter cloacae infection, 2 cases of MRSA, 2 cases of acinetobacter baumannii infection, 2 cases of klebsiella pneumoniae infection, 1 case in enterococcus faecium and pseudomonas aeruginosa and staphylococcus haemolyticus, respectively. There were still 7 patients with negative bacterial culture. 41 cases retained their implant, whereas 1 staphylococcus aureus infection patient had the implants removed because of loosening during debridement. Nevertheless, primary wound healing was found in all patients, and stitches were removed 2 to 3 weeks after debridement. Infections were effectively controlled with no recurrence of infection during the follow-up. The average erythrocyte sedimentation rate was (65.76±20.08) mm/h preoperative, (41.43±14.65) mm/h 1 month postoperative, (10.81±2.72) mm/h 6 months postoperative, and (8.10±5.46) mm/h 12 months postoperative, respectively, the differences were statistically significant. The average C reactive protein was (40.55±16.91) mg/L preoperative, (6.50±2.46) mg/L 1 month postoperative, (4.31±1.26) mg/L 6 months postoperative, and (3.83±1.50) mg/L 12 months postoperative, respectively, the differences were statistically significant. The average procalcitonin was (0.47±0.28) ng/ml preoperative, (0.08±0.06) ng/ml 1 month postoperative, (0.06±0.03) ng/ml 6 months postoperative, and (0.05±0.00) ng/ml 12 months postoperative, respectively, and the differences were statistically significant. Conclusion A timely diagnosis, aggressive and meticulous debridement, high vacuum closed-suction drain, routine and adequate use of antibacterial agents are keys to successfully resolving infection and maintaining implant retention in the treatment of delayed deep infection after spinal fixation. Key words: Infection; Reoperation; Debridement