1. We investigated the effect of the nitric oxide (NO) synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME) and the peptidase alpha-chymotrypsin on non-adrenergic, non-cholinergic (NANC neural) bronchoconstriction induced by electrical stimulation of the vagus nerves and by capsaicin in anaesthetized guinea-pigs in vivo using pulmonary insufflation pressure (PIP) as an index of bronchial tone. We also investigated the contribution of soluble guanylyl cyclase (SGC) to NANC neural relaxant mechanisms. 2. In the presence of atropine and propranolol, electrical stimulation of the vagus nerves induced a frequency-dependent increase in PIP above baseline of 67% at 2.5 Hz, of 128% at 5 Hz and of 230% at 10 Hz. L-NAME (1-50 mg kg-1, i.v.), at doses inducing increases in systemic blood pressure, dose-relatedly potentiated NANC bronchoconstriction. At 10 mg kg-1 i.v., L-NAME significantly (P < 0.05) potentiated NANC bronchoconstriction by a further 106% at 2.5 Hz and a further 147% at 5 Hz but did not potentiate the increase in PIP at 10 Hz. L-NAME did not induce bronchoconstriction in sham-stimulated control animals. D-NAME did not potentiate NANC bronchoconstriction. Raising systemic blood pressure with phenylephrine did not potentiate vagally-induced bronchoconstriction (2.5 Hz). 3. The NO precursor L-arginine, but not D-arginine, (100 mg kg-1, i.v.) significantly reversed the potentiation by L-NAME of NANC bronchoconstriction. L-Arginine alone significantly inhibited neurogenic bronchoconstriction at 10 Hz (by 74%); the inhibition of 25% at 2.5 Hz was not significant. 4. L-NAME did not significantly affect the increases in PIP induced by intravenous substance P. neurokinin A (NKA) or capsaicin. 5. The inhibitor of SGC, methylene blue (10 mg kg', i.v.) potentiated (by 110-140%) NANC neural bronchoconstriction induced by lower frequencies of nerve stimulation and reversed the reduction in PIP induced by the SGC activator, sodium nitroprusside (SNP, 1.05 mg kg- 1, i.v.). SNP significantly (P <0.05) reduced by 65% the bronchoconstriction induced by nerve stimulation at 10 Hz. Methylene blue did not effect baseline PIP in sham-stimulated controls. The airway effects of methylene blue and SNP were not associated with their cardiovascular effects. 6. a-Chymotrypsin (2 units kg-', i.v.) significantly potentiated vagally-induced bronchoconstriction by a further 63% at 2.5 Hz, by a further 95.6% at 5 Hz but did not potentiate the increase in PIP at 10 Hz. alpha-Chymotrypsin also potentiated (by 116%) capsaicin-induced bronchoconstriction. Vasoactive intestinal peptide (VIP, 10 ig kg-' i.v. infused over min) significantly reduced by 70% the increase in PIP induced by NKA (0.1 .Lmol kg-' i.v., infused over 30 s). 7. The combination of a-chymotrypsin (2 units kg-', i.v.) and L-NAME (5 mg kg-', i.v.) significantly potentiated NANC bronchoconstriction by a further 304% at 2.5 Hz, an increase in PIP which was greater than that induced by either a-chymotrypsin or L-NAME alone (P <0.05). 8. We conclude that endogenous NO and a bronchodilator peptide, possibly VIP, released in association with nerve stimulation, as well as activation of soluble guanylyl cyclase, regulate the magnitude of NANC neurogenic bronchoconstriction in guinea-pigs in vivo.