1 The gastric adaptation re¯ex is activated by the release of non-adrenergic, non-cholinergic (NANC) inhibitory transmitters, including nitric oxide (NO) and vasoactive intestinal polypeptide (VIP). The role of NO in this re¯ex is not disputed, but some investigators suggest that NO synthesis is stimulated by VIP in post-junctional cells or in nerve terminals. We investigated whether the e ects of these transmitters are mediated by independent pathways in the canine gastric fundus. 2 VIP and NO produced concentration-dependent relaxation of the canine fundus. N o -nitro-Larginine (L-NNA) reduced relaxation induced by electrical ®eld stimulation (EFS; 0.5 ± 8 Hz), but had no e ect on responses to exogenous VIP and sodium nitroprusside (SNP, 10 mM). 3 Oxyhaemoglobin reduced relaxations produced by EFS and SNP. Oxyhaemoglobin also reduced relaxation responses to low concentrations of VIP (510 nM), but these e ects were non-speci®c and mimicked by methaemoglobin which had no e ect on nitrergic responses. 4 A blocker of guanylyl cyclase, 1H-[1,2,4]oxidiazolo [4,3,-a]quinoxalin-1-one, (ODQ) inhibited responses to EFS, SNP and DETA/NONOate (an NO×donor), but had no e ect on responses to VIP. cis-N-(2-phenylcyclopentil)-azacyclotridec-1en-2-amine monohydrochloride (MDL 12,330A), a blocker of adenylyl cyclase, reduced responses to EFS, VIP and forskolin, but did not a ect responses to SNP. 5 Levels of cyclic GMP were enhanced by the NO donor S-nitroso-n-acetylpenicillamine (SNAP) but were una ected by VIP (1 mM). The increase in cyclic GMP in response to SNAP was blocked by ODQ. 6 The results suggest that at least two transmitters, possibly NO and VIP, mediate relaxation responses in the canine fundus. NO and VIP mediate responses via cyclic GMP-and cyclic AMPdependent mechanisms, respectively. No evidence was found for a serial cascade in which VIP is coupled to NO-dependent responses.