The benefit of noninvasive pressure support ventilation (NIPSV) in avoiding the need for endotracheal intubation and reducing morbidity and mortality associated with endotracheal intubation was evaluated in 41 patients who presented with acute respiratory failure not related to chronic obstructive pulmonary disease (COPD). Patients were randomly assigned to receive conventional therapy (n = 20) or conventional therapy plus NIPSV (n = 21). NIPSV was delivered to the patient by a face mask connected to a ventilator (Puritan-Bennett 7200a) set in inspiratory pressure support (IPS) mode. The mean levels of IPS, positive end-expiratory pressure (PEEP), and fraction of inspired oxygen (FIO2) were respectively 15 +/- 3 cm H2O, 4 +/- 2 cm H2O, and 57 +/- 22%. The rate of endotracheal intubation (62 vs 70%, p = 0.88), the length of ICU stay (17 +/- 19 days vs 25 +/- 23 days, p = 0.16), and the mortality rate (33 vs 50%, p = 0.46) were not different between patients treated with NIPSV and those treated conventionally. Post hoc analysis suggested that in patients with PaCO2 > 45 mm Hg (n = 17), NIPSV was associated with a reduction in the rate of endotracheal intubation (36 vs 100%, p = 0.02), in the length of ICU stay (13 +/- 15 days vs 32 +/- 30 days, p = 0.04), and in the mortality rate (9 vs 66%, p = 0.06). We conclude that NIPSV is of no benefit when used systematically in all forms of acute respiratory failure not related to COPD. A subgroup of patients, characterized by acute ventilatory failure and hypercapnia, may potentially benefit from this therapy and further studies are needed to focus on this aspect.