1. The chemosensitivity of cough receptors stimulated by inhalation of aqueous aerosols was evaluated in 21 normal volunteers in three experiments. 2. The pH of isotonic saline was altered using small amounts of phosphate or glycine buffers to produce solutions with a pH range of 2.6-10.0. These solutions were nebulized ultrasonically and breathed for 1 min periods by seven subjects in random order and on separate days. Cough frequency during each 1 min inhalation was recorded. Only the two solutions of extreme pH (2.6 and 10.0) caused cough. 3. The effect of altering the osmolarity of the inhaled aerosol on cough was assessed using D-glucose over a range of 77-1232 mosmol/l. Saline solutions over the same range of osmolarity were also tested. The pH of D-glucose was raised to match that of saline by adding small amounts of sodium hydroxide. All solutions were nebulized and inhaled by seven subjects as described above for 1 min periods during which cough frequency was recorded. Forced expired volume in 1 s was recorded after each inhalation and did not alter in any subject by more than 10%. Subjects coughed when inhaling all the D-glucose solutions over the whole range of osmolarity. Cough occurred with saline solutions only at low chloride concentration and at the highest concentration. 4. In order to clarify whether the response to hypertonic saline was due to the high ionic content of the solutions or to its hypertonicity, two other solutions were tested. These were an isotonic and a hypertonic mixture of D-glucose and saline, containing 'normal' (150 mmol/l) ionic content.(ABSTRACT TRUNCATED AT 250 WORDS)