Postoperative infection remains a troublesome but not uncommon complication after spinal surgery. Most previous reports, however, are small or involve cases with more than one surgeon often at different institutions. This study represents a single surgeon's 9-year experience with postoperative infection at one institution. The authors describe the features of wound infection after spinal surgery with reference to diagnosis, microbiology, and treatment and they describe a protocol for effective management of postoperative spinal wound infection. The records of the senior author (F.P.C.) during a 9-year period for cases of postoperative wound infection were reviewed. Of 2,391 operative procedures, 46 cases of wound infection were identified, yielding an overall infection rate of 1.9%. Patients' preoperative risk factors, original diagnosis prompting the surgery, onset of infection, presentation, treatment, and outcome were analyzed. The mean age of the 23 men and 23 women was 57.2 years. The preoperative diagnoses included lumbar degenerative scoliosis or spinal stenosis in 28 cases, disk prolapse in 8 cases, metastatic disease in 4 cases, degenerative disk disease in 1 case, and a group of 5 miscellaneous cases. Seventeen (37%) of the patients underwent at least one previous spinal surgery at the same site. Twenty-three patients had a fusion, of whom 22 also had instrumentation. Forty-three (93%) of the patients had significant wound drainage after an average of 15 days (range, 5–80 days). The other three patients were examined approximately 2 years after the surgery. Fourteen of the patients also had pyrexia (temperature >37.5°C) at presentation. Staphylococcus aureus alone was cultured in 29 patients, whereas another six patients had a different single organism. In nine patients, more than one organism was cultured during their hospital stay. Surgical treatment included primary closure in only seven patients, with most undergoing wound drainage and debridement followed by delayed closure. Instruments were removed in the three patients with late presentation who had solid fusion at operation. Viable bone graft and instrumentation were left in situ in all patients who were seen before fusion. All wounds healed without sequelae, except for three that required flap closure. Pseudarthrosis was noted in three patients after more than 1 year of follow-up in this series. Postoperative spinal wound infection is a potentially devastating problem. In this series, infection was more common in patients undergoing fusion with instrumentation and in patients with cancer metastatic to the spine. An aggressive surgical approach, including repeated debridement followed by delayed closure, is justified. Instrumentation may be safely left in situ to provide stability for fusion.