Background and objective: Induction of anaesthesia promotes collapse of dependent lung regions in both obese and nonobese patients. We hypothesized that end-expiratory lung volume (EELV) may be more sensitive than oxygenation to evaluate the effects of positive end-expiratory pressure (PEEP) after anaesthesia induction. Methods: Forty patients (20 nonobese patients and 20 obese patients) were prospectively studied. After anaesthesia induction, PEEP was adjusted in a stepwise fashion [zero end-expiratory pressure (ZEEP), PEEP 5 cmH2O and PEEP 10 cmH2O]. At each step, we measured EELV, static elastance, gas exchange and dead space. Other than changing PEEP, respiratory settings were kept constant throughout. Results: Anaesthesia induction and ZEEP both lowered EELV by 39% in nonobese patients and 59% in obese patients (both P < 0.05), as well as oxygenation (P < 0.05). Compared with ZEEP, in nonobese patients, PEEP 5 cmH2O and PEEP 10 cmH2O improved EELV (+15 and +40%, respectively, P < 0.01) and elastance but not oxygenation. In obese patients, PEEP 10 cmH2O also improved EELV (49% vs. ZEEP and 30% vs. PEEP 5 cmH2O, P < 0.01), elastance and dead-space fraction, with no effect on oxygenation. PEEP-induced changes of EELV correlated with changes of elastance (r = 0.46, P = 0.003), but not with oxygenation. Conclusion: After induction of anaesthesia, mechanical ventilation with ZEEP is associated with a profound reduction in EELV. PEEP improves efficiently EELV and respiratory mechanics, with no major effect on oxygenation. EELV may be a useful indicator to guide PEEP setting in the operating room.