A study was carried out to find out whether the use of a hygrophobic filter (Pall, Ultipor) or of a heated humidifier (Dräger, Aquapor) during surgery had any effect on a patient's intraoperative core temperature and thermal balance. Seventy-five ASA I or II patients scheduled for gynaecological surgery were randomly assigned to three groups: group A (n = 25), where no warming device was used; and two groups (n = 25 for each) where inhaled gases were humidified and heated with either a hygrophobic filter set up between the endotracheal tube and the Y-piece (group B) or a heated humidifier set to 100% saturation at a temperature of 41.5 degrees C (group C). The patients were all anaesthetised with the same technique (thiopentone 5 mg.kg-1, dextromoramide 0.03 mg.kg-1 and 0.1 mg.kg-1 of either pancuronium or vecuronium, followed by enflurane with nitrous oxide in oxygen); the perfused fluids were not heated. Room, tympanic, rectal, oesophageal and four skin (thorax, arm, leg, thigh) temperatures were measured with calibrated Exacon thermistances, on arrival in the operating theatre, during induction, every ten minutes for two hours, and then every twenty minutes for two hours more. Ramanathan's and Burton's formulae were used to calculate mean skin temperature and heat loss respectively. In the recovery room, patients were warmed up with an electric blanket. Shivering was ranked from "0" to "+ +". There were no differences between groups as far as age, drug doses, perfusion volumes and room temperature were concerned.(ABSTRACT TRUNCATED AT 250 WORDS)