Inhaled frusemide protects asthmatic airways against a wide variety of bronchoconstrictor stimuli by unknown mechanisms. To investigate whether inhaled loop diuretics modulate baseline bronchial responsiveness, a randomized, double-blind, placebo-controlled study was conducted to test the ability of frusemide (40 mg) and bumetanide (2 mg) to displace concentration-response curves with methacholine in 14 healthy volunteers. In addition, separate randomized, double-blind studies were carried out to evaluate the effects of oral flurbiprofen, a potent cyclo-oxygenase inhibitor, on the protective action of frusemide against methacholine-induced bronchoconstriction. Inhaled loop diuretics significantly increased the provocative concentration of methacholine causing a 15% decrease in forced expiratory volume in one second (PC15FEV1) from the geometric mean (range) value of 58.6 (9.2-233) mg.ml-1 after placebo administration, to 129 (13.8-505) and to 106 (6.6-510) mg.ml-1 after administration of frusemide and bumetanide, respectively. Similar results were obtained when data from partial flow-volume curves were used for analysis. In the eight subjects studied, pretreatment with oral placebo and inhaled frusemide reduced airway responsiveness to methacholine, with a geometric mean (range) PC15FEV1 value of 116 (25.4-405) mg.ml-1, and premedication with oral flurbiprofen abolished this protective effect, the geometric mean (range) PC15FEV1 methacholine being reduced to a value of 50.3 (16.6-189) mg.ml-1. In addition, oral flurbiprofen alone failed to alter airway responsiveness to methacholine. In view of these findings, it is suggested that bronchoprotective prostaglandins may mediate the effects of loop diuretics against methacholine-induced bronchoconstriction in man.