BACKGROUND: Rapid elimination of nitrous oxide from the lungs at the end of inhalational anesthesia dilutes alveolar oxygen, producing "diffusion hypoxia." A similar dilutional effect on accompanying volatile anesthetic agent has not been evaluated and may impact the speed of emergence. METHODS: Twenty patients undergoing surgery were randomly assigned to receive an anesthetic maintenance gas mixture of sevoflurane adjusted to bispectral index, in air-oxygen (control group) versus a 2:1 mixture of nitrous oxide-oxygen (nitrous oxide group). After surgery, baseline arterial and tidal gas samples were taken. Patients were ventilated with oxygen, and arterial and tidal gas sampling was repeated at 2 and 5 min. Arterial sampling was repeated 30 min after surgery. Sevoflurane partial pressure was measured in blood by the double headspace equilibration technique and in tidal gas using a calibrated infrared gas analyzer. Time to eye opening and time extubation were recorded. The primary endpoint was the reduction in sevoflurane partial pressures in blood at 2 and 5 min. RESULTS: Relative to baseline, arterial sevoflurane partial pressure was 39% higher at 5 min in the control group (P < 0.04) versus the nitrous oxide group. At 30 min the difference was not statistically significant. Time to eye opening (8.7 vs. 10.1 min) and time to extubation (11.0 vs.13.2 min) were shorter in the nitrous oxide group versus the control group (P < 0.04). CONCLUSIONS: Elimination of nitrous oxide at the end of anesthesia produces a clinically significant acceleration in the reduction of concentrations of the accompanying volatile agents, contributing to the speed of emergence observed after inhalational nitrous oxide anesthetic.