BACKGROUND Surgical site infection (SSI) is a potentially morbid and costly complication following major colorectal resection. In recent years, there has been growing attention placed on the accurate identification and monitoring of such surgical complications and their costs, measured in terms of increased morbidity to patients and increased financial costs to society. We hypothesize that incisional SSIs following elective colorectal resection are more frequent than is generally reported in the literature, that they can be predicated by measurable perioperative factors, and that they carry substantial morbidity and cost. METHODS Over a 2-year period at a university hospital, data on all elective colorectal resections performed by a single surgeon were retrospectively collected. The outcome of interest was a diagnosis of incisional SSI as defined by the Center of Disease Control and Prevention. Variables associated with infection, as identified in the literature or by experts, were collected and analyzed for their association with incisional SSI development in this patient cohort. Multivariate analysis by stepwise logistic regression was then performed on those variables associated with incisional SSI by univariate analysis to determine their prognostic significance. The incidence of SSI in this study was compared with the rates of incisional SSI in this patient population reported in the literature, predicted by a nationally based system monitoring nosocomial infection, and described in a prospectively acquired intradepartmental surgical infection data base at our institution. RESULTS One hundred seventy-six patients undergoing elective colorectal resection were identified for evaluation. The mean patient age was 62 +/- 1.2 years, and 54% were men. Preoperative diagnoses included colorectal cancer (57%), inflammatory bowel disease (20%), diverticulitis (10%), and benign polyp disease (5%). SSIs were identified in 45 patients (26%). Twenty-two (49%) SSIs were detected in the outpatient setting following discharge. Of all preoperative and perioperative variables measured, increasing patient body mass index and intraoperative hypotension independently predicted incisional SSI. Although we could not measure statistically increased length of hospital stay associated with SSI, a representative population of patients with SSI accumulated a mean of $6200/patient of home health expenses related to wound care. Our rates of SSI were substantially higher than that reported generally in the literature, predicted by the National Nosocomial Infection System, or described by our own institutional surgical infection data base. CONCLUSIONS The incidence of incisional SSI in patients undergoing elective colorectal resection in our cohort was substantially higher than generally reported in the literature, the NNIS or predicted by an institutional surgical infection complication registry. Although some of these differences may be attributable to patient population differences, we believe these discrepancies highlight the potential limitations of systematic outcomes measurement tools which are independent of the primary clinical care team. Accurate surgical complication documentation by the primary clinical team is critical to identify the true frequency and etiology of surgical complications such as incisional SSI, to rationally approach their reduction and decrease their associated costs to patients and the health care system.