The objective of this study was to evaluate a non-volitional measurement to assess diaphragmatic function in intubated and mechanically ventilated patients in a prospective pilot interventional clinical trial. The study was conducted in an 18-bed postoperative intensive care unit based at a university hospital. Patients were prospectively assigned to two groups. Group 1 consisted of eight patients with ventilator weaning failure. Group 2 consisted of eight intubated and ventilated patients who were studied shortly after major surgery and were successfully extubated there-after The twitch pressure response after cervical magnetic stimulation of the phrenic nerves was measured at the endotracheal tube at different PEEP levels. In group 2 the twitch transdiaphragmatic pressure, defined as the difference between twitch gastric and twitch oesophageal pressure was also evaluated. In group 1 the mean twitch pressure at the endotracheal tube on PEEP 0, 5 and 10 cmH2O was 5.2, 4.5 and 2.6 cmH2O: In group 2 this was significantly higher (15.1 cmH2O on PEEP 0 and 12.2 cmH2O on PEEP 5). A good correlation was found between twitch diaphragmatic pressure and twitch pressure at the endotracheal tube (r2 = 0.96) and between twitch oesophageal pressure and twitch pressure at the endotracheal tube (r2 = 0.98). Patients with weaning failure have significantly lower twitch pressure at the endotracheal tube suggesting diaphragmatic dysfunction. Twitch pressure at the endotracheal tube may be a useful parameter to screen for diaphragmatic dysfunction in intubated critically ill patients. Further studies are needed to confirm these preliminary findings.