A multicenter study was undertaken to analyze postoperative wound infections after posterior spinal instrumentation and fusion. The infection rate of these procedures has been documented in multiple reports. From these results, a classification scheme was developed that can guide therapy and determine the populations at risk. The patients were categorized according to two parameters, the first being the severity or type of infection, and the second being the host response or physiologic classification of the patient. This classification scheme is based on the clinical staging system for adult osteomyelitis developed by Cierny. The severity of infection is divided into three groups. Group 1 is a single-organism infection, either superficial or deep. Group 2 is a multiple-organism, deep infection. Group 3 is multiple organisms with myonecrosis. The host response, likewise, is divided into three classes. Class A is a host with normal systemic defenses, metabolic capabilities, and vascularity. Class B patients demonstrate local or multiple systemic diseases, including cigarette smoking. Class C requires an immunocompromised or severely malnourished host. Our data have demonstrated that single organisms, Group 1, generally can be dealt with by single irrigation and debridement, and closure over suction drainage tubes without the use of an inflow-irrigation system. The Group 2 patients, with multiple organisms and deep infection, required an average of three irrigation debridements. They have a higher percentage of successful closures with closed inflow-outflow suction irrigation systems when compared to simple suction drainage systems without constant inflow irrigation. Multiple-organism infections with myonecrosis, Group 3, are exceedingly difficult to manage, and portend a poor outcome. Patients without normal host defenses, Classes B and C, are at high risk for developing postoperative wound infection. Specifically, this study demonstrated that cigarette smoking may be a significant risk factor.