We compared the effects of pretreatment with saline, ipratropium bromide, and disodium cromoglycate (DSCG) on bronchoconstriction induced by methoxamine‐an alpha‐adrenoceptor agonist, in asthmatic subjects. All 12 patients bronchoconstricted in response to methoxamine after saline. The PD20 (the dose of methoxamine causing a 20% fall in forced expiratory volume in 1 s [FEV1]) ranged from 0.3‐18 mumol. Ipratropium bromide (200 micrograms administered by aerosol) significantly inhibited (P less than 0.05) the response to methoxamine in all patients without producing significant changes in the mean baseline lung function. The mean PD20 for methoxamine after saline was 6.8 mumol and 95% confidence limits (CL) were 3.6, 12.7 mumol. The mean PD20 for methoxamine after ipratropium bromide was 35.4 (95% CL 28.8, 43.6) mumol. DSCG also produced significant (P less than 0.05) shifts to the right in the methoxamine dose response curves, but did not affect resting airway calibre as measured by the FEV1. The mean PD20 for methoxamine increased from 3.3 mumol (95% CL 1.1, 10.0 mumol) after saline to 25.1 mumol (95% CL 14.1, 44.6) after DSCG pretreatment. These findings suggest that alpha‐adrenoceptors in the airways of asthmatic subjects may be located at sites other than smooth muscle‐possibly on mast cells but more likely on nerve endings and/or parasympathetic ganglia.