Adrenaline may have beneficial effects in asthma in addition to a direct beta-adrenoceptor mediated bronchodilatation, such as alpha-receptor mediated reduction in microvascular leakage and oedema, and inhibition of bronchoconstrictor neural pathways. We have compared the bronchodilator effect of nebulised adrenaline (1 mg) with nebulised salbutamol (2.5 mg) in patients with acute severe asthma. Eighteen patients admitted with acute asthma (mean peak expiratory flow 22% predicted) were randomised to receive either adrenaline or salbutamol in a double-blind fashion and, after 15 min, were changed to the alternative nebulisation to determine if there was any additional bronchodilation. There were no differences between the increase in PEF after adrenaline (mean +/- SEM increase 99 +/- 20.5 L/min) or after salbutamol (119 +/- 22.7 L/min). Heart rate fell after each nebuliser and there was no difference between treatments. PaO2 rose after adrenaline (0.5 +/- 0.15 kPa), but fell after salbutamol (-0.2 +/- 0.11 kPa). These results suggest that nebulised adrenaline is as effective as a nebulised beta-agonist in acute asthma and is without significant side-effects. The theoretical advantages conferred by alpha-agonist activity do not produce any additional bronchodilation but may prevent any fall in PaO2 due to ventilation-perfusion mismatching.