Background: Sevoflurane expenditure, inspired gas humidity, temperature, soda lime temperature, and compounds A and B were measured during high and low fresh gas flow anaesthesia in paediatric patients. Methods: Sixty ASA 1 or 2 paediatric patients were randomly allocated to two groups: low-flow circle anaesthesia (LFA) patient group (n = 30) and high-flow circle anaesthesia (HFA) patient group (n = 30). initial fresh gas flow (FGF) was 4 l . min(-1) of nitrous oxide and 2 l . min(-1) of oxygen in both groups. This FGF of 6 l . min(-1) was maintained in the HFA group. After 10 min of HFA, the FGF was reduced to 600 ml . min(-1) (nitrous oxide and oxygen 300 ml . min(-1) each) in the LFA group. Results: Sevoflurane expenditure during LFA was about 1/7 of that during HFA (3.3 +/- 0.2 ml . h(-1) . vol.%(-1) compared to 22.8 +/- 0.6 ml . h(-1) . vol.%(-1), mean +/- SEM, respectively). Absolute humidity in the LFA patients was 4 times higher than that in the HFA patients (22.8 +/- 2.4 g . m(-3) and 5.6 +/- 3.4 g . m(-3) respectively). There was no significant difference in the inspiratory gas temperature between the LFA (28.5 +/- 0.6 degrees C) and HFA (26.9 +/- 1.3 degrees C) groups. There was significant difference in the mean highest soda lime temperature between the LFA (35.5 +/- 1.2 degrees C) and HFA (28.7 +/- 1.2 degrees C) groups. The mean highest concentration of compound A was 12.2 +/- 3.8 ppm in the LFA group. The mean highest concentration of compound B was less than 1 ppm. Compounds A and B were below detectable level in the HFA group. Conclusion: In conclusion, sevoflurane used for paediatric patients in a circle system with a fresh gas flow of 0.6 l . min(-1) resulted in a significantly reduced sevoflurane expenditure, higher inspired absolute humidity, but not temperature, compared to a fresh gas flow of 6 l . min(-1). Low levels of compounds A and B were detected.