One hundred and eleven children with acute asthma were studied to compare delivery of terbutaline by either a metered dose inhaler (MDI) with a valved holding chamber or a nebuliser driven by air. Eligible patients were randomised; the MDI group received three puffs (0.75 mg) of terbutaline and the nebuliser group received 2 ml (5.0 mg) terbutaline solution diluted with 2 ml 0.9% saline for inhalation over 10 minutes. Patients were evaluated by spirometry, pulse oximetry, and clinical severity scoring system at baseline and again 15 minutes after the beginning of treatment. The baseline data of the two groups were not significantly different. All parameters of spirometry, except the peak expiratory flow (PEF) for the nebuliser group, and clinical severity score for both groups significantly improved after terbutaline treatment. Compared with the nebuliser group, the MDI group after treatment had better mean (SD) oxygen saturation (SaO2; 96.82 (1.63)% v 95.44 (1.88)%), frequency of oxygen desaturation (23.2% v 47.3%), absolute increase of PEF (32.6 (37.7) l/min v 10.2 (34.7) l/min), and SaO2 (0.54 (1.64)% v -0.47 (1.84)%). There was also a mean (SD) per cent increase of forced expiratory volume in one second (22.9 (21.0)% v 15.4 (16.1)%), PEF (27.7 (38.4)%) v 7.7 (25.1)%), and SaO2 (0.58 (1.72)% v -0.47 (1.93)%). In conclusion, aerosol treatment by MDI (with a valved holding chamber) in this study proved to be superior to nebuliser treatment in terms of SaO2 and some measurements of spirometry. Respiratory therapists working with children with severe asthma should be aware of the possibility of oxygen desaturation, especially when using room air as the driving gas for nebulisation.