89 subjects with stable asthma took part in a double-blind, placebo-controlled, randomized, crossover study of the effects of regular versus on-demand inhaled bronchodilator therapy. The subjects inhaled fenoterol or placebo by a dry powder delivery system for 24 weeks. Control of asthma was judged by daily morning and evening peak expiratory flow rates, symptom diaries, use of additional inhaled bronchodilator, and requirement for short courses of prednisone. Of 64 subjects who completed the trial, 57 showed a clear difference in degree of control of asthma between the fenoterol and placebo periods: in 17 (30% [95% confidence interval 18.4-43.4%]) asthma was better controlled during regular inhaled bronchodilator treatment, whereas in 40 (70% [56.6-81.6%]) control was better during placebo treatment with bronchodilator for symptom relief only. Mean airway responsiveness to methacholine increased slightly during the fenoterol period. The adverse effect of regular bronchodilator inhalation occurred not only among subjects who used a bronchodilator as sole treatment (2 were better and 10 were worse during regular bronchodilator treatment) but also among those who took inhaled corticosteroids (14 were better and 29 were worse). Thus, regular inhalation of a beta-sympathomimetic agent was associated with deterioration of asthma control in the majority of subjects. The trends to use of regular, higher doses or longer-acting inhaled beta-sympathomimetic treatment may be an important causal factor in the worldwide increase in morbidity from asthma.