The management of facial fractures in the polytrauma patient requires the coordination of multiple surgical disciplines to optimize the functional and cosmetic outcome while minimizing overall morbidity and mortality. Although the plastic surgery literature historically advocates the early repair of facial fractures, the risk of general anesthesia in patients with associated injuries sometimes makes early repair unsafe. We compared early operative repair versus delayed operative repair of facial fractures in multitrauma patients. We specifically examined wound infection, overall complication rate, total length of hospital stay, days in the Intensive Care Unit (ICU), and days on the ventilator in the two groups. A 5-year (1991-1996) retrospective study of multitrauma patients with associated facial fractures was undertaken at an urban community hospital. We had a total of 82 patients, who were divided into three groups. Thirty-three patients did not have operative repair of the facial fractures during the initial admission and were followed by the plastic surgery service on an outpatient basis. These patients will not be discussed further. Seven patients underwent early operative repair, which was defined as repair within 48 hours of admission (group I). Forty-two patients had delayed operative repair, defined as repair more than 48 hours after admission (group II). The reasons for delayed repair included: excessive soft tissue swelling (16), intracranial injuries (12), unstable medical condition (8), and coordination of procedures with other services (6). Of the 49 patients who underwent operative repair, 43 were involved in motor vehicle accidents, 3 were injured by a fall from a height, 2 were involved in auto-pedestrian accidents, and 1 was a victim of assault. Forty-eight of the 49 patients were initially admitted to the ICU. Cumulative associated injuries were as follows: closed head injury (38), extremity fracture (21), blunt chest injuries (11), intra-abdominal injuries (5), vertebral column injuries (7), and ocular injuries (2). The average Injury Severity Score for Group I was 17.3 and for Group II, 18.1. In group I, there were no deaths, there were no wound infections, and the complication rate was 14.3 per cent. The average total number of days spent on the ventilator was 3.0, the average total number of days spent in the ICU was 5.0, and the average total hospital stay was 16.0 days. In group II, there were no deaths, the wound infection rate was 5 per cent, and the overall complication rate was 21 per cent. The average total number of days spent on the ventilator was 3.3, the average total number of days spent in the ICU was 5.8, and the average total days in the hospital was 14.8. Our results indicate that in a similar cohort of multitrauma patients, delayed repair did not increase length of ICU stay or hospital stay. The wound infection rate was negligible, and the complication rate was similar in the two groups. We conclude that a delay in repair of facial fractures in the critically ill patient has an acceptably low complication rate and may be advantageous in decreasing operative risk and minimizing cost by coordinating multiple procedures with various surgical subspecialties.