In 1970, the Centers for Disease Control and Prevention (CDC) established the National Nosocomial Infection Surveillance System to assist institutions with infection surveillance, data collection, and processing. This facilitates interinstitutional comparison for nosocomial infection rates. Nosocomial infection rates in the surgical intensive care unit have been shown to be different from the medical intensive care unit. Whether there exists a difference in infection rates between trauma and surgical patients in the intensive care unit has not been established. Our objective was to determine whether there is a difference in rates of nosocomial infections between trauma and surgical patients in the surgical intensive care unit. From January 1995 through December 1997, we reviewed 3715 admissions to the surgical intensive care unit and separated them into trauma (1272) or surgical (2443) cases. We documented all nosocomial pneumonias, urinary tract infections, bloodstream infections, and surgical site infections. From these data we determined infection rates per 100 admissions. We also identified all device-related nosocomial infections and calculated infection rate by current CDC standards using number of device infections divided by number of device-days times 1000. We found that the overall trauma patient infection rate was 11.64 per cent compared with 6.43 per cent for surgical patients (P<.001). Using conventional infection rate criteria, trauma patients had higher frequency in the rate of ventilator-associated pneumonia (6.13% vs. 2.50%; P<0.001), urinary tract infection (2.36 versus 1.76; P<0.2), and bloodstream infection (2.52% versus 1.27%; P<0.01). However, when using the CDC guidelines, which correct for the number of device-days for infections, only the difference in rate of pneumonia between the two groups reached statistical significance (23.9 rate for trauma patients vs. 16.7 for the surgery group; P<0.005). We conclude that trauma patients are at higher risk for nosocomial infections than routine surgical patients. Because of this difference, centers should collect and report data separately for trauma and surgical patients in the intensive care unit. Specific attention should be focused on the causes and prevention of increased rates of nosocomial pneumonia in trauma patients.