This study compared the rates of acute respiratory failure, reintubation, length of intensive care stay and mortality in patients in whom the non-invasive mechanical ventilation (NIMV) was applied instead of the routine venturi face mask (VM) application after a successful weaning. Following the approval of the hospital ethics committee, 62 patients who were under mechanical ventilation for at least 48hours were scheduled for this study. 12 patients were excluded because of the weaning failure during T-tube trial. The patients who had optimum weaning criteria after the T-tube trial of 30minutes were extubated. The patients were kept on VM for 1hour to observe the hemodynamic and respiratory stability. The group of 50 patients who were successful to wean randomly allocated to have either VM (n=25), or NIV (n=25). Systolic arterial pressure (SAP), heart rate (HR), respiratory rate (RR), PaO2, PCO2, and pH values were recorded. The number of patients who developed respiratory failure in the NIV group was significantly less than VM group of patients (3 reintubation vs. 14 NIV+5 reintubation in the VM group). The length of stay in the ICU was also significantly shorter in NIV group (5.2±4.9 vs. 16.7±7.7 days). The ratio of the respiratory failure and the length of stay in the ICU were lower when non-invasive mechanical ventilation was used after extubation even if the patient is regarded as 'successfully weaned'. We recommend the use of NIMV in such patients to avoid unexpected ventilator failure.