To examine the sensitivity of different parts of the human respiratory tract to a tussive and bronchoconstrictor stimulus, randomized, standardized single breaths of capsaicin aerosols were inhaled by nine healthy, nonsmoking subjects. A small droplet aerosol (3.2-mu-m MMD) was inhaled slowly (0.25 L/s), and a large droplet aerosol (5.2-mu-m MMD) was inhaled rapidly (1.0 L/s) to optimize differences in deposition, which were assessed after inhalation of Tc-99m-DTPA aerosols with similar characteristics. Both capsaicin aerosols (zem to 256-mu-M) produced a concentration-dependent cough response. The geometric means (95% Cl) for the concentrations causing two coughs (cough threshold) were 2.3-mu-M (1.1 to 4.9) and 8.7-mu-M (3.1 to 24.5) (p < 0.02), respectively, with the small and large droplet aerosols. The concentrations causing five coughs were 5.5-mu-M (3.0 to 10.0) with the small droplet aerosol and 29.5-mu-M (8.3 to 104.7) with the large droplet aerosol (p < 0.02). In contrast, FEV1, measured 2 min after the largest concentration of capsaicin, was not significantly altered by any of the two capsaicin aerosols. In each subject, a similar dose was deposited in the larynx with both aerosols, whereas the dose deposited in intrapulmonary airways was 2.3 times larger with the small droplet aerosol. This study confirmed that slow inhalation of a small droplet aerosol produced a more peripheral airway deposition than did rapid Inhalation of a large droplet aerosol. The small droplet aerosol was four times more potent, and capsaicin-sensitive sensory neurons mediating cough seem, therefore, to be present in human intrapulmonary airways. These nerve endings appear to be more sensitive to a chemical stimulus than those in the larynx. Irrespective of site of deposition, the capsaicin aerosols did not significantly alter airway tone, supporting the view that cough and reflex bronchoconstriction are mediated via separate afferent pathways.