OBJECTIVE To evaluate the safety and efficiency of the use of the laryngeal mask airway (LMA) during percutaneous dilatational tracheostomy under bronchoscopic guidance comparing with the ventilation via endotracheal tube (ET). METHODS Prospective, randomized clinical trial in the eight-bed general intensive care unit of a university hospital. METHODS 60 consecutive adult critically ill patients who required elective tracheostomy for a period of 12 months. METHODS Patients were randomly assigned to ventilated via LMA ( n=30 patients), and to ventilated via ET ( n=30). RESULTS Blood samples for arterial blood gas analyses were taken before the procedure (first value) and just before the insertion of tracheostomy tube (second value). There was no significant difference in pH, PaO2, or PaCO2 between groups before the procedure. The operating time was significantly shorter in LMA group (4.5+/-0.8 min versus 5.9+/-1.4 min). Although the second PaCO2 values were higher than the first in both groups, the rise in was significantly higher in ET group (6.8+/-3.5 mmHg vs. 4.5+/-2.4 mmHg). Hypercarbia was noted in 10 patients (38.5%) in the LMA group and 17 (56.7%) in the ET group. The decrease in pH related to hypercarbia was noted in both groups, but it was more significant in the ET group ( p<0.05). CONCLUSIONS LMA is an effective and successful ventilatory device during percutaneous dilatational tracheostomy. It improves visualization of the trachea and larynx during fiberoptic-assisted percutaneous dilatational tracheostomy and prevents the difficulties associated with the use of ET such as cuff puncture, tube transection by the needle, and accidental extubation. The use of a bronchoscope and the puncture of the ET cuff cause major increases in PaCO2.