OBJECTIVE: To determine the incidence of and risk factors for surgical-site infections (SSIs) after abdominal surgery. DESIGN: A cohort study was conducted from January to June 1998. CDC criteria for SSI and the NNIS System risk index were used. SETTING: A tertiary-care hospital in Peru. PATIENTS: Adult patients undergoing abdominal surgery who consented were enrolled and observed until 30 days after surgery. Patients who had undergone surgery at another hospital or who died or were transferred to another hospital within 24 hours after surgery were excluded. RESULTS: Four hundred sixty-eight patients were enrolled. Their mean age was 37.2 years. One hundred twenty-five patients developed SSIs, 18% of which were identified after discharge. The overall incidence rate (IR) was 26.7%. The IR was 13.9% for clean, 15.9% for clean-contaminated, 13.5% for contaminated, and 47.2% for dirty interventions. The IR was 3.6% for NNIS System risk index 0 and 60% for index 3. Risk factors for SSI on logistic regression analysis were dirty or infected wound (RR, 3.8; CI95, 1.7-8.4), drain use longer than 9 days (RR, 6.0; CI95, 2.5-12.5), and length of surgery greater than the 75th percentile (RR, 2.1; CI95, 1.0-4.4). Patients with SSI had a longer hospital stay than did non-infected patients (14.0 vs 6.1 days; P<.001). CONCLUSIONS: SSI is a major problem in this hospital, which has a higher IR (especially for clean interventions) than those of developed countries. In developing countries, prevention of SSI should include active surveillance and interventions targeting modifiable risk factors.