We have documented the physiologic effects of fiberoptic bronchoscopy with bronchoalveolar lavage (BAL) and endobronchial biopsy performed under local anesthesia in 20 asthmatic subjects, 8 healthy nonatopic control subjects, and 8 atopic nonasthmatic subjects. Premedication consisted of nebulized albuterol (2.5 mg; except for the study of atopic nonasthmatic subjects), ipratropium bromide (500 micrograms), and intramuscular atropine (0.6 mg). Intravenous midazolam was given for mild sedation, and oxygen was delivered via a nasal cannula. FEV1 was measured before and after premedication, immediately postbronchoscopy, and after 2 h recovery. There was a significant fall in mean (+/- SD) FEV1 immediately postbronchoscopy in both the asthmatic (26.2 +/- 16.7%; p less than 0.001) and normal (9 +/- 4.7%, p less than 0.05) groups, which in the asthmatic subjects correlated inversely with the concentration of methacholine provoking a 20% fall in FEV1 (PC20) measured 5 days prebronchoscopy (r = -0.74, p less than 0.001) but not with symptom scores, albuterol use, or peak expiratory flow (PEF) variation recorded during 2 wk before the investigation. There was significant arterial hemoglobin O2 desaturation during biopsy in the asthmatic subjects (median 3%, range -1 to 17% fall from baseline; p less than 0.01), which was not related to any of the measured indices of asthma severity. PC20, measured 5 days before and 5 days after bronchoscopy in the asthmatic subjects and 2 days before and 1 day after bronchoscopy in the atopic nonasthmatic subjects was not significantly affected by the procedure. We conclude that fiberoptic bronchoscopy with BAL and endobronchial biopsy can be conducted safely in asthmatic subjects, but requires caution in those with very responsive airways.(ABSTRACT TRUNCATED AT 250 WORDS)