Gastric emptying of orange-flavoured glucose (group I), low-fat milk (group II) and breast milk (group III) was evaluated in 45 ASA grade I children of < or = 5 years of age by using real-time ultrasonography and residual gastric volume and pH was then measured. In 15 more children, residual gastric volume and pH was measured after a midnight fast (group IV). Mean (SD) gastric emptying time in group I was 1.53 (0.25) h (range 1.00-1.75), group II 2.32 (0.31) h (range 1.75-2.75) and group III 2.43 (0.27) h (range 2.00-2.75). According to Robert and Shirley's criteria, no children of group I and II were found to be 'at risk' at 2 h and 3 h, respectively, but 13.3% of group III children were labelled as 'at risk' at 3 h. The incidence of 'at risk' children in group IV was 33.3%. It was concluded that 3% fat milk or 17.5% glucose in a volume of 10 ml.kg-1 (maximum volume of 100 ml) can be given in children safely 3 h and 2 h, respectively, before anaesthesia. More real-time studies are required on breast milk to establish guidelines for its potential use as a pre-operative feed 3 h before anaesthesia.