A bench study followed by a clinical trial were performed to evaluate the mechanical characteristics of five (commercially available) expiratory valves used for home ventilators, as well as the potential clinical impact of differences between these valves. In the in vitro study, expiratory valve resistance was evaluated under unvarying conditions, whereas dynamic behaviour was evaluated by calculating the imposed expiratory work of breathing during a simulated breath generated by a lung model. Differences in resistance and imposed expiratory work of up to twofold and 150%, respectively, were found across valves. We then conducted a randomized crossover clinical study to compare the effects of the least resistive (Bennett) and most resistive expiratory valves (Peters) in 10 intubated patients receiving pressure support ventilation. There were no significant differences regarding blood gases or respiratory parameters except for the oesophageal pressure-time product (PTPoes), which was significantly increased by the Peters valve (236+/-113 cmH2O x s x min(-1) versus 194+/-90 cmH2O x s x min(-1)). An analysis of individual responses found that the Peters valve induced substantial increases in intrinsic positive end-expiratory pressure (PEEP), PTPoes, and expiratory activity in those patients with the greatest ventilatory demand. In conclusion, differences between home expiratory valve resistances may have a clinically relevant impact on the respiratory effort of patients with a high ventilatory demand.