In 15 asthmatic children post-exercise bronchospasm was partially inhibited by placebo and by aerosolized atropine sulphate (1 mg) compared with no treatment, not significantly inhibited by atropine alone compared with placebo, partially blocked by disodium cromoglycate (20 mg) and by isoproterenol (0.625 mg) and completely blocked by the combination of isoproterenol and atropine. Pre-treatment with isoproterenol or atropine resulted in post-exercise values for specific conductance which were significantly greater than those following disodium cromoglycate by virtue of the bronchodilator effect of these drugs independent of or in addition to any specific inhibition of exercise-induced asthma. These results suggest that the bronchoconstrictor response to exercise is partially influenced by suggestion but is influenced to a significantly greater degree by mediator release and beta-adrenergic mechanisms and that bronchodilator drugs have therapeutic advantages over an inhibitor of mediator release in the prevention of exercise-induced asthma.