We have studied 28 patients mechanically ventilated for acute respiratory failure at different levels of externally applied positive end-expiratory pressure (PEEPe). We describe and compare a computed method of measuring "effective" elastance of the total respiratory system (Ers,eff) with the static values of elastance of the total respiratory system (Ers,st), obtained with the end-inflation occlusion technique. Ers,eff was computed by an original device (Heres, R.P.A., Belgium), also the effective resistance of the total respiratory system was calculated. At zero end-expiratory pressure set by the ventilator (ZEEP). Ers,eff averaged 29.5 +/- 13.5 cm H2O x L-1 while Ers,st non-corrected for intrinsic PEEP (PEEPi) averaged 36.4 +/- 15.1 cm H2O x L-1 and Ers,st corrected for PEEPi averaged 28.2 +/- 13.4 cm H2O x L-1. The small difference between Ers,eff and Ers,st corrected for PEEPi was statistically significant and these two values were highly correlated (r = 0.98). This significant difference disappeared rapidly with PEEPe and probably reflects a frequency-dependance due to pendelluft. We also observed that PEEPi was present in 21 of 27 patients at ZEEP. Our results also indicate that low levels of PEEP may improve Ers in hyperinflated COPD patients, without inducing further hyperinflation. In conclusion, values of Ers,eff are very similar to static values corrected for PEEPi and permit an accurate and rapid approach to the management of ventilated patients.