Sixteen children, aged 2 to 5 years and ranked ASA 1, were included in this study assessing gastro-oesophageal reflux occurring under halothane anaesthesia, before and during, caudal anaesthesia. They were scheduled for surgery below the umbilicus lasting 1 to 5 h. After premedication with oral hydroxyzine (2 mg.kg-1) and intravenous atropine (10 micrograms.kg-1), induction was carried out with 3% halothane. A gastro-oesophageal pH probe was inserted via the nose after calibration at 37 degrees C. A neutral pH for the oesophageal electrode and an acid pH for the gastric one demonstrated the correct position of the probe. The pH was then registered every 4 s. The probe was left in situ until the patient left the recovery room. The caudal anaesthesia catheter was then inserted with the patient lying on his left side. Caudal anaesthesia was began with 2.5 mg.kg-1 of plain bupivacaine and 5 mg.kg-1 of plain lidocaine. When the patient was lying supine again, narcosis was maintained with 0.5% halothane and 50% nitrous oxide. A dose of 1.5 mg.kg-1 of bupivacaine was injected every 30 to 45 min. None of the children displayed any respiratory signs (coughing, dyspnoea, bronchospasm, cyanosis) during the combined anaesthetic. Two episodes of asymptomatic gastro-oesophageal reflux were revealed by this method, one lasting 7 minutes and occurring during insertion of the caudal catheter, and the other, lasting 4 minutes, during recovery. There were no pulmonary sequels. There was excellent respiratory and haemodynamic stability throughout. The two episodes seemed to have been triggered off by rapid displacement of the patient and too deep an anaesthetic.(ABSTRACT TRUNCATED AT 250 WORDS)