Our institution used an experimental protocol for the use of inhaled amphotericin B as a prophylactic measure to prevent fungal disease in severely immunocompromised patients. We did a prospective study of the physiologic effects of amphotericin B administration. We looked specifically at oxygen saturation levels, peak flow values, and symptoms of patients given amphotericin B. We collected data on a series of 18 patients and of 132 amphotericin B administrations. Four (22%) of the patients stopped treatments because of nausea and vomiting which were believed to be due to the inhaled amphotericin B. For the remaining patients, no treatment was stopped because of symptoms or physiologic changes caused by amphotericin B, although there were 9 instances of clinically significant bronchospasm as defined by a drop in peak flow of 20% or more, 9 clinically relevant increases in cough, and 3 clinically relevant increases in dyspnea. Forty-eight percent of the clinically relevant changes occurred in patient 8. Another 16% occurred in asthmatic subjects who were significantly more likely (p = 0.03) to experience a 20% or more drop in peak flow than were patients without asthma. The physiologic profile of the response to inhaled amphotericin B is acceptable.