<h3>Objective</h3> To describe the frequency of aerosol pentamidine-induced bronchoconstriction, its relationship to nonspecific airway responsiveness, and its response to preventive therapy using salbutamol, ipratropium bromide, or sodium cromoglycate. <h3>Methods</h3> Consecutive HIV-infected individuals starting prophylactic AP were eligible if they had not been previously treated with this agent. Simple spirometry was performed before and 10 min after a single 60-mg dose given through an ultrasonic nebulizer. Methacholine challenge was performed in all subjects 24 h to four days after the initial AP dose. Subjects with a change in FEV<sub>1</sub> (ΔFEV<sub>1</sub>)≥10 percent decrease after the initial AP dose were restudied on three separate occasions (=24 hours apart) after premedication with two puffs of salbutamol (200 μg), ipratropium bromide (40 μg), or sodium cromoglycate (2 mg), in random order. <h3>Results</h3> Fifty-three subjects were studied. The median ΔFEV<sub>1</sub> after a single dose of AP was — 7.0 percent (range: -47 percent, 1.8 percent). The ΔFEV<sub>1</sub> following AP was only partially predicted by the degree of nonspecific bronchial responsiveness as measured by a standard methacholine challenge. Age, current smoking, history of asthma, baseline FEV<sub>1</sub>, or a prior episode of PCP failed to predict the ΔFEV<sub>1</sub> following AP. Eighteen subjects (34 percent) had a ΔFEV<sub>1</sub> ≥10 percent decrease (median: —17.0 percent). In these subjects, after premedication with salbutamol, ipratropium bromide, and sodium cromoglycate, the median ΔFEV<sub>1</sub> was 1.0, 0.8, and —9.6 percent, respectively. <i>Conclusion:</i> Aerosol pentamidine produced a decrease in FEV<sub>1</sub>≥10 percent in 34 percent of subjects. This was not accurately predicted by the methacholine response. The bronchoconstriction induced by AP was effectively prevented by either salbutamol or ipratropium, whereas cromoglycate was only partially effective. <i>(Chest 1991; 100:624-27)</i>