Inhaled furosemide prevents bronchoconstriction induced by nebulized distilled water, exercise, and antigen challenge. We examined the effect of furosemide on cough induced by low chloride content solutions and by capsaicin in double-blind, placebo-controlled studies. A group of eight nonsmoking normal subjects was given furosemide (3.75 mg/ml inhaled for 8 min) and placebo (saline) immediately before consecutive 1-min inhalations of four isosmolar solutions with decreasing chloride content every 5 min from an ultrasonic nebulizer. Decreasing concentrations of chloride induced dose-related coughing, which was inhibited by furosemide. Thus, chloride-free solution induced 13.1 +/- 1.6 coughs after placebo and 8.4 +/- 1.9 coughs after furosemide (p less than 0.005). In a separate study, six of the same normal subjects were given inhaled furosemide or placebo before inhaling one breath of capsaicin solution given in three consecutive increasing concentrations. Capsaicin induced dose-related coughing, which was not inhibited by furosemide. Thus, after placebo the highest concentration of capsaicin induced 20.8 +/- 1.8 coughs and after furosemide, 21.5 +/- 2.7 coughs. We conclude that furosemide may act by inhibiting the cough reflex indirectly, perhaps by changing local chloride ions within the vicinity of epithelial cough receptors.