Intrinsic positive end-expiratory pressure (PEEPi) has usually been interpreted as suggesting dynamic hyperinflation, but expiratory muscle activity may also increase end-expiratory alveolar pressure without any additional increase in end-expiratory lung volume. The aim of this study was to assess the influence of expiratory muscle activity, which increases abdominal pressure during expiration and is followed by a sudden drop at end-expiration, on PEEPi measurement in mechanically ventilated patients. We studied eight tracheally intubated patients breathing in an assisted mode in whom expiratory muscle activity was present. PEEPi was measured from the fluctuations of esophageal pressure (Pes) while continuous recording of gastric pressure (Pga) and of changes in abdominal cross-sectional area assessed expiratory muscle activity. PEEPi was also measured by the airway occlusion method in one patient, and diaphragmatic electromyographic activity was recorded to determine the timing of inspiratory muscle activity in two patients. Varying the level of ventilatory support (pressure support level, peak flow rate, or PEEP level) induced increases in measured PEEPi from 6.7 +/- 3.4 to 13.2 +/- 5.9 cm H2O. Concomitantly, the expiratory rise in Pga increased from 3.1 +/- 2.7 to 8.6 +/- 5.0 cm H2O, and the abrupt decay in Pga observed at the end of expiration increased from 4.2 +/- 3.7 to 10.6 +/- 6.1 cm H2O. The drop in Pga and the drop in Pes at end-expiration were synchronous, and these changes, together with electromyographic measurements, were consistent with a concomitant relaxation of the expiratory muscles and activation of the inspiratory muscles.(ABSTRACT TRUNCATED AT 250 WORDS)