140 children of 184 with acute asthma entered a randomised double-blind trial of oral prednisolone (n = 67) compared with placebo (n = 73) administered soon after admission. The dose of prednisolone was 30 mg in children under 5, otherwise 60 mg. All children also received salbutamol. All had moderate or severe dyspnoea. Initial evaluation was similar for both groups. On reassessment after a few hours 20 children in the prednisolone group were fit for discharge compared with only 2 in the placebo group. There were no early reattendances. Children remaining in hospital had a shorter median duration of stay and were less likely to require further steroid therapy if they had initially received prednisolone. In acute asthma the prompt use of a single dose of oral prednisolone can reduce morbidity and the need for hospital care.