The exact mechanism of adenosine-induced bronchoconstriction in patients with asthma is unknown. Adenosine contraction of guinea pig trachea was antagonized by inhibitors of cyclooxygenase. The aim of this study was to investigate the effect of indomethacin (100 mg/day) on adenosine-induced bronchoconstriction in 14 asymptomatic patients with asthma. Airway response was evaluated as FEV1, and adenosine was administered as an aerosol diluted in 0.9% saline to produce a concentration range of 0.125 to 4 mg/ml. The dose of adenosine producing a 20% change in FEV1 (PD20) was calculated from the individual semilogarithmic dose-response curve; the results of PD20 were converted to log values for statistical analysis (Student's paired t test). The study was performed on 3 separate days. On the first day, the adenosine challenge was performed, and on subsequent days patients were pretreated with either placebo or indomethacin in a randomized, double-blind manner. Inhaled adenosine caused bronchoconstriction with a geometric mean PD20 of 0.71 mg (95% confidence limits, 0.44 to 1.16). After placebo, a geometric mean PD20 of 0.91 mg (95% confidence limits, 0.53 to 1.58) was obtained. Indomethacin pretreatment decreased adenosine hyperresponsiveness and shifted the dose-response curves of adenosine challenge to the right with a geometric mean PD20 of 1.28 mg (95% confidence limits, 0.64 to 2.56). The effect of indomethacin on adenosine bronchoconstriction (p less than 0.01 versus baseline; p less than 0.05 versus placebo) suggests an indirect mechanism of adenosine on inducing release of arachidonic acid derivatives. Inflammatory mediators inhibited by indomethacin may be involved in adenosine bronchoconstriction, even if this mechanism is not relevant.(ABSTRACT TRUNCATED AT 250 WORDS)