Objective The purpose of this study was to compare the three techniques: medium-flow, low-flow and low-flow with BIS monitoring, for sevoflurane anaesthesia in terms of consumption of sevoflurane, recovery from anaesthesia, awakening time and side effects. Methods Ninety-six ASA Ⅰ - Ⅱ aged 27-51 yr undergoing elective surgery on low abdominal or low extremities under general anaesthesia were randomly divided into three groups: group A medium flow (FGF 1000 ml?min-1 ), group B low flow (FGF 500 ml?min-1 ) and group C low flow ( FGF 500 ml? min-1 ) with BIS monitoring. Sevoflurane was delivered into the circuit system from a Komesarroff vaporizer placed in-circle on the inspiration limb. In groups A and B the concentration of sevoflurane delivered was adjusted according to clinical signs of anesthesia, while in group C according to the BIS value (at 46 ?10). Before induction of anesthesia the patient was denitrogenated for 3 min with high flow rate of oxygen (6 L ?min-1 ). Anesthesia was induced with midazolam 0.03 mg?kg-1 , fentanyl 1 ?g?kg-1 , propofol 2 mg?kg-1 and vecuronium 0.1 mg?kg-1. After intubation, the patient was mechanically ventilated and PaCO2 was maintained at 35 - 45 mm Hg. Anesthesia was maintained with sevoflurane. The consumption of sevoflurane was calculated from deduction of the volume of sevoflurane left in the vaporizer from 30ml of sevoflurane added initially into the vaporizer. The duration from termination of sevoflurane administration to eye-opening and orientation and the incidence of nausea and vomiting were recorded. Results During surgery the end-tidal sevoflurane concentrations were maintained at (1.40?0.20) MAC (in group A), (1.10? 0.20) MAC (in group B) and (0.80?0.20) MAC (in group C) respectively. The volume of sevoflurane consumed was (13.3 ? 1.6) ml?h-1 (group A), (9.6 ?1.5 ) ml ? h ( group B) and (7.5?1.8)ml?h-1( group C) respectively. The time to regain consciousness were (14.3?3.3) min (group A), (10.5 ? 2.8) min (group B) and (7.5?2.6) min (group C). The times to full orientation were (24.5?6.1) min (group A), (17.4?5.5) min (group B) and (12.7 ? 4.8) min (group C). The incidence of nausea and vomiting was 14.5 % ? 2.6 % (group A), 10.1 % ?2.3 % (group B) and 7.5 % ?2.1 % (group C) . Conclusion Low-flow closed circuit anaesthesia combined with BIS monitoring has the advantages of least sevoflurane consumed, fastest recovery and least incidence of nausea and vomiting and is the best technique for sevoflurane anaesthesia.