The neurotransmitters responsible for neurogenic airway relaxation are still unknown. We investigated the effects of N ω -nitro-L-arginine methylester (L-NAME) on nonadrenergic and noncholinergic (NANC) relaxation evoked by electrical stimulation of vagus nerves in vivo and in vitro in cat.To that end, we measured pulmonary resistance during vagal nerve stimulation (VS) in vivo, and isometric tension of small bronchi (1-3 mm outer diameter) during electrical field stimulation (EFS) in vitro. During infusion of 5-hydroxytryptamine (5-HT), VS transiently decreased total pulmonary resistance in the presence of atropine and propranolol, with peak relaxation at several seconds after the VS and a gradual return to baseline within 2-3 min. L-NAME abolished the initial peak relaxation and reduced the peak amplitude, but did not affect the duration of the NANC relaxation. In small bronchi obtained from control cats, EFS evoked a biphasic NANC relaxation, comprising an initial fast component followed by a second slow component, and L-NAME (10 -5 M) selectively abolished the first component without affecting the second. Whilst in the small bronchi obtained from L-NAME pretreated cats, EFS elicited only the slow component of NANC relaxation, which was insensitive to L-NAME but sensitive to tetrodotoxin.These results indicate that nonadrenergic noncholinergic relaxation induced by vagal nerve stimulation during infusion of 5-hydroxytryptamine can be classified into two components, and that at least two neurotransmitters, including nitric oxide, are involved in the relaxation. Eur Respir J 1997; 10: 314- Neurally-mediated relaxation of airway smooth muscle in various animal species, including man, is largely nonadrenergic and noncholinergic (NANC) (see, for example, [1]). Although the neurotransmitter(s) responsible for NANC relaxation in the airways have not been conclusively identified, nitric oxide (NO) and vasoactive intestinal polypeptide (VIP) have emerged as strong candidates [2][3][4][5][6][7].In cat airway, we have shown that NANC relaxation can be classified into two components by thresholds for activation or by sensitivity to N ω -nitro-L-arginine methylester (L-NAME), and that the pattern of L-NAME-sensitive and -insensitive components differs in central and peripheral airways [6,7]. Specifically, electrical field stimulation (EFS) elicited monophasic relaxation in the trachea, confirming previous observations [5,8], but biphasic NANC relaxations, comprising an initial fast component followed by a second slow component in the bronchioles. L-NAME selectively abolished the first component of NANC relaxation without affecting the second in bronchioles, whilst, in the trachea, L-NAME completely suppressed the monophasic NANC relaxation after single or short repetitions (<5) of 1 ms pulse stimuli; however, after more stimuli (>10) of 1 or 4 ms duration, suppression of NANC relaxation was incomplete (to 40-50% of the control value) [7]. VIP antagonists partially suppressed the L-NAME-resistant NANC relaxation....