1. The effects of inhibitors of nitric oxide synthase and local anaesthetics were studied on changes in human nasal airway patency and albumin extravasation in response to bradykinin and histamine, in vivo. 2. Compared with the action of the vasoconstrictor, ephedrine, 2.5 mumol, NG-nitro-L-arginine methyl ester (L-NAME), 1 mumol alone, did not change the resting value of the minimal cross-sectional area (A min) of the human nasal airway. L-NAME, 0.1 to 10 mumol, produced a dose-related inhibition of the reduction in A min caused by bradykinin, 300 micrograms. NG-monomethyl-L-arginine (L-NMMA), 1 mumol, similarly reduced the effect of bradykinin, 300 micrograms, on A min, but NG-nitro-D-arginine methyl ester (D-NAME), had no effect. L-NAME, 0.1 to 10 mumol, or L-NMMA, 10 mumol, failed to inhibit the effect of histamine, 300 micrograms on A min. 3. The inhibition by L-NAME, 1 mumol of the action of bradykinin, 300 micrograms on A min was maximal between 15 and 30 min after pretreatment with L-NAME. 4. L-NAME, 1 and 10 mumol, inhibited the extravasation of albumin into the nasal cavity induced by bradykinin, 300 micrograms, and also by histamine, 300 micrograms. D-NAME, 1 and 10 mumol had no effect on the extravasation of albumin in response to bradykinin or histamine. 5. L-Arginine, 30 mumol, reversed the effect of L-NAME, 1 mumol, on the bradykinin- and histamine-induced albumin extravasation into the nasal airway. 6. Local anaesthesia of the nasal airway with lignocaine, 10 mg, or benzocaine, 10 mg, failed to inhibit the reduction in A min or the albumin extravasation induced by either bradykinin, 300 micrograms, and histamine, 300 micrograms. 7. We conclude that the extravasation of plasma albumin caused by bradykinin and by histamine involves the generation of nitric oxide. The nasal blockage induced by bradykinin involves nitric oxide generation but the nasal blockage induced by histamine does not.