BACKGROUND: Mechanical ventilation is the defining event of intensive care unit management. To reduce use, a literature-based protocol was introduced to facilitate weaning. The effect of protocol-driven ventilator weaning on ventilator use, ventilator-associated pneumonia (VAP), and intensive care unit (ICU) length of stay (LOS) is described in a survey of 2 years' activity in a multidisciplinary surgical ICU. METHODS: Data were gathered from April to September 2000 and from April to September 2002 before and after introduction of nurse/therapist-driven weaning. VAP was identified by chest radiography, clinical presentation, Gram's stains, and cultures from tracheal aspirates or bronchoalveolar lavage. Infection control practitioners diagnosed VAP. Failed extubation was defined as reintubation within 72 hours. RESULTS: Overall, there was a 2:1 ratio of male patients to female patients. The total number of patients and days of mechanical ventilation increased, but the use ratio (ventilator days/ICU days) fell from 0.47 to 0.33. Patients failing extubation fell from 43 (in 2000) to 25 (in 2002). From these patients, 17 cases of VAP occurred in 2000 and 5 in 2002. Mean age (40 years), Injury Severity Score (24), and ICU LOS (5.7-7.4 days; p = not significant) were unchanged in injured patients. ICU discharge was frequently delayed because of the need for subsequent respiratory care. CONCLUSION: Protocol-driven weaning reduces use of mechanical ventilation and VAP. Injured and general surgical patients show reduction in complications, but shorter ICU LOS depends on resources elsewhere in the health care system.