Of 357 children with acute diarrhoea admitted to the City Hospital, Edinburgh, over a 12-month period, only 5 (1.4%) required IV infusions. Three hundred and nineteen were treated with oral rehydration (OR). Of these 269 were studied in detail and 43% had signs of dehydration, but in none of them was it severe. There were no fatalities. Patients were randomly allocated to treatment with one of three OR solutions in a double blind trial. The solutions had sodium concentrations of 35, 50 and 90 mmol (mEq)/l, and dextrose of 200 (36 g/l), 111 (20 g/l) and 110 (19.8 g/l) mmol/l, respectively. Hypernatraemia was not a clinical problem and only 5 children (2%) were biochemically hypernatraemic on admission. Treatment did not cause clinical hypernatraemia. At the second assessment only 3 children were biochemically hypernatraemic, one from each treatment group, and no one had clinical signs. All three solutions were safe and effective in the relatively mildly dehydrated patients currently seen in the UK.