The helmet, a transparent latex-free polyvinyl chloride cylinder linked by a metallic ring to a soft collar that seals the helmet around the neck, has been recently proposed as an effective alternative to conventional face mask to deliver pressure support ventilation (PSV) during noninvasive ventilation in patients with acute respiratory failure. We tested the hypothesis that mechanical characteristics of the helmet (large internal volume and high compliance) might impair patient-ventilator interactions compared with standard face mask. Breathing pattern, CO(2) clearance, indexes of inspiratory muscle effort and patient-ventilator asynchrony, and dyspnea were measured at different levels of PSV delivered by face mask and helmet in six healthy volunteers before (load-off) and after (load-on) application of a linear resistor. During load-off, no differences in breathing pattern and inspiratory muscle effort were found. During load-on, the use of helmet to deliver pressure support increased inspiratory muscle effort and patient-ventilator asynchrony, worsened CO(2) clearance, and increased dyspnea compared with standard face mask. Autocycled breaths accounted for 12 and 25% of the total minute ventilation and for 10 and 23% of the total inspiratory muscle effort during mask and helmet PSV, respectively. We conclude that PSV delivered by helmet interface is less effective in unloading inspiratory muscles compared with PSV delivered by standard face mask. Other ventilatory assist modes should be tested to exploit to the most the potential benefits offered by the helmet.