OBJECTIVE An endotracheal cuff pressure of 20-30 cmH(2)O is recommended. Underinflation and overinflation are associated with complications such as aspiration and tracheal wall damage. The aim of this study was to identify prevalence of, and risk factors for, endotracheal cuff underinflation and overinflation. METHODS Prospective observational cohort study. All critically ill patients intubated with a high-volume lowpressure endotracheal tube were eligible. After manual adjustment of cuff pressure at 25 cmH(2)O, continuous recording of cuff pressure and airway pressure was performed for 8 h. Underinflation and overinflation of the endotracheal cuff were defined as cuff pressure less than 20 cmH(2)O and more than 30 cmH(2)O, respectively. In all patients, the time spent with normal cuff pressure or with underinflation or overinflation of the endotracheal cuff was measured. Univariate and multivariate analyses were used to determine risk factors for cuff underinflation and overinflation. RESULTS Eight hundred and eight hours of cuff pressure recordings were analysed in 101 patients. Eighteen per cent of study patients spent 100% of recording time with normal (20-30 cmH(2)O) cuff pressure. Fifty-four per cent of study patients developed cuff underinflation, 73% developed cuff overinflation, and 44% developed both. Thirty- three per cent of study patients developed underinflation or overinflation for more than 30 min. Absence of sedation [odds ratio (95% confidence interval)=2.51 (1-6), P=0.03] and duration of prior intubation [1.16 (1.04-1.29), P<0.01] were independently associated with cuff underinflation. No risk factor for overinflation could be determined. The percentage of time spent with underinflation significantly (P<0.01) increased during the recording period. CONCLUSIONS Variations in endotracheal cuff pressure are common in ICU patients. Duration of prior intubation and absence of sedation are independently associated with increased risk for cuff underinflation.