Background: Noninvasive positive pressure ventilation (NPPV) using facial or nasal mask have been widely used for several years in stable patients with chronic neuromuscular disease or central alveolar hypoventilation, and recently have been tried in patients with acute respiratory failure. In a few studies, NPPV was also used to rescue the patients with post-extubation respiratory failure. However, yet it has not been adopted as a weaning method in patients on long-term mechanical ventilation. So we performed this prospective clinical study to evaluate the usefulness of NPPV as a weaning method after removing endotracheal tube intentionally in patients on long-term mechanical ventilation. Method: Twelve patients who had been on invasive mechanical ventilation over 10 days were enrolled and 14 trials of NPPV were done. All had failed at least one weaning trial and showed ventilator dependence(pressure support requirement between 8-15cm $H_2O$, and PEEP requirement between 5-10cm $H_2O$), so tracheostomy was being considered. After removing the endotracheal tube, NPPV was applied using facial mask. Respiratory rate, arterial blood gas, pressure support level, and PEEP level were monitored just before intended extubation, at 30 minutes, 1 to 6, 6 to 12, 12 to 24 hours, 2nd day, and 3rd day following initiation of NPPV, and just before weaning from NPPV. The successful weaning was defined as spontaneous breathing off the ventilator for 48 hours or longer without respiratory distress. Results: The weaning through NPPV after intended extubation was successful in 7(50%) of 14 trials, and tracheostomy could be avoided in them. There were no differences in age, sex, APACHE III score, duration of invasive mechanical ventilation, baseline respiratory rate, $PaCO_2$ $PaO_2/FiO_2$, and ventilatory requirement(PS and PEEP) between the success and failure groups. In the success group, respiratory rate, pH, $PaCO_2$, and $PaO_2/FiO_2$ were not different between invasive MV and NPPV period. But in the failure group, pH decreased after 30 minutes of NPPV initiation compared with that of invasive MV($7.40\pm0.08$ vs. $7.34\pm0.06$, p<0.05). The causes of failure were worsening of ABG(n=3), retained tracheal secretion(n=2), mask intolerance(n=1), and flail chest(n=1). Conclusion: NPPV may be worth trying as a bridge method in weaning patients on long-term invasive mechanical ventilation.