The relationship between frequency of ventilator circuit changes and risk of ventilator-associated pneumonia was studied using 2 independent approaches. The first was an in-use aerosol contamination study with patients on 8-hour, 16-hour, or 24-hour ventilator changing schedules. The second approach was a study comparing the incidence of pneumonia in patients on ventilators for 2 one-year periods when the ventilator circuit changing time differed. In-use aerosol sampling of 513 ventilator treatment periods showed bacterial contamination greater than 100 organisms per aerosol in 1.8 per cent of 8-hour cycles, in 2.5 per cent of 16-hour cycles, and in 5.4 per cent of 24-hour cycles. These differences were not significant. The incidence of ventilator-associated pneumonia was the same for one-year periods when ventilator circuits were changed either every 8 or every 24 hours. It was concluded that changing ventilator circuits every 24 hours provides adequate protection from ventilator-associated pneumonia.