BACKGROUND: Many of the complications of mechanical ventilation are related to inappropriate endotracheal tube (ETT) cuff pressure. The aim of the current study was to evaluate the effectiveness of automatic cuff pressure closed-loop control in patients under prolonged intubation, where presence of carbon dioxide (CO2) in the subglottic space is used as an indicator for leaks. The primary outcome of the study is leakage around the cuff quantified using the area under the curve (AUC) of CO2 leakage over time. METHODS: This was a multicenter, prospective, randomized controlled, noninferiority trial including intensive care unit patients. All patients were intubated with the AnapnoGuard ETT, which has an extra lumen used to monitor CO2 levels in the subglottic space. The study group was connected to the AnapnoGuard system operating with cuff control adjusted automatically based on subglottic CO2 (automatic group). The control group was connected to the AnapnoGuard system, while cuff pressure was managed manually using a manometer 3 times/d (manual group). The system recorded around cuff CO2 leakage in both groups. RESULTS: Seventy-two patients were recruited and 64 included in the final analysis. The mean hourly around cuff CO2 leak (mm Hg AUC/h) was 0.22 ± 0.32 in the manual group and 0.09 ± 0.04 in the automatic group (P =.01) where the lower bound of the 1-sided 95% confidence interval was 0.05, demonstrating noninferiority (>-0.033). Additionally, the 2-sided 95% confidence interval was 0.010 to 0.196, showing superiority (>0.0) as well. Significant CO2 leakage (CO2 >2 mm Hg) was 0.027 ± 0.057 (mm Hg AUC/h) in the automatic group versus 0.296 ± 0.784 (mm Hg AUC/h) in the manual group (P =.025). In addition, cuff pressures were in the predefined safety range 97.6% of the time in the automatic group compared to 48.2% in the automatic group (P <.001). CONCLUSIONS: This study shows that the automatic cuff pressure group is not only noninferior but also superior compared to the manual cuff pressure group. Thus, the use of automatic cuff pressure control based on subglottic measurements of CO2 levels is an effective method for ETT cuff pressure optimization. The method is safe and can be easily utilized with any intubated patient.