We tested the hypothesis that pressure-support ventilation (PSV) allows a reduction in emergence time and laryngeal mask airway (LMA) removal time after general anesthesia compared to volume-controlled mechanical ventilation (CMV). Because spontaneous breathing (SB) is often used with LMA under general anesthesia, patients were allocated randomly to three groups (CMV, SB and PSV). Thirty-six consecutive ASA I-II patients scheduled for knee arthroscopic surgery under general anesthesia with a LMA and breathing throughout the ventilator circuit were included. Hemodynamic and ventilatory variables were recorded before and 10-min after general anesthesia-induction, at the surgical incision, at the end of anaesthetic drugs infusion and when the patient was totally awake (which defines emergence time). LMA removal time, drug consumption were recorded at the end of the surgical procedure. Leak fraction around the LMA was also evaluated. LMA removal time was significantly higher in the CMV-group (18¡6 min) compared to both SB (8¡4 min) and PSV (7¡4 min, P,0.05) groups as well as for emergence time: CMV-group (32¡12 min), SB (17¡7 min) and PSV (13¡6 min, P,0.05) groups. Total propofol consumption was significantly lower in the PSV-group (610¡180 mg) than in both CMV (852¡330 mg) and SB (734¡246 mg, P,0.05) groups. Air leaks around the LMA was significantly higher in the CMV-group than in the SB and PSV groups (16% vs 3% and 7%, all P,0.05). In conclusion, in knee arthroscopic surgery, in comparison to CMV, PSV use during general anesthesia in unparalyzed patients decreases LMA removal time, propofol consumption and leaks around LMA while improving ventilatory variables without adverse effects.