This study tested the hypothesis that the final mediator of nonadrenergic, noncholinergic (NANC) inhibitory junction potentials (ijps) and associated relaxation responses was nitric oxide (NO) or a related substance and not vasoactive intestinal polypeptide (VIP). We used opossum esophagus body circular muscle and canine intestine circular muscle. In both these tissues, ijps had reversal potentials near the potassium equilibrium potential, (EK); in esophagus the ijps were apamin insensitive, but in the intestine they were partially apamin sensitive. Nω-Nitro-L-arginine methyl ester (NAME) (10−5 to 5 × 10−4 M) abolished ijps in both tissues, an effect overcome by 10−3 M L-arginine but not D-arginine. NAME increased input resistance of esophagus tissues in the double sucrose gap but caused no significant depolarization in the sucrose gap or in studies with microelectrodes. Contractions and basal tension were increased in both tissues by NAME. The apamin sensitive and insensitive ijp components in canine muscle were both abolished by NAME, but the time course of this abolition was different for the two components. Methylene blue (10–50 μM) with variable rapidity and extent inhibited ijps in both tissues, but L-arginine could not overcome this effect. Methylene blue, like NAME, did not depolarize detectably but enhanced the contractile activity. VIP (10−6 M) had very small effects in both tissues, little or no hyperpolarization and increased input resistance in esophagus, these effects were not changed by NAME, and VIP did not affect ijps. We conclude that NO may be the final mediator of NANC-initiated inhibitory junction potentials in gastrointestinal circular smooth muscle.Key words: nitric oxide, nonadrenergic noncholinergic inhibitory mediator, vasoactive intestinal polypeptide, esophagus, intestine.