Since nitric oxide has been found to control the function of many organs of the body by the non-adrenergic, non-cholinergic branch of the autonomic nervous system, we hypothesized that it might play a role in salivary secretion. Therefore, we investigated the distribution of nitric oxide synthase (NOS) throughout the submaxillary gland and also studied the ability of inhibitors of NOS to interfere with salivation induced by a cholinergic agonist, metacholine, and by a polypeptide, substance P. The secretory responses were determined in rats anesthetized with chlorolose following intravenous injection of the various pharmacological agents. There was no basal flow of saliva and dose-response curves were obtained by sequential intravenous injection of increasing doses of the drugs. Then, in the same animal, the same dose-response curves were performed in the presence of NOS inhibitors. L-Nitro-arginine-methyl-ester (L-NAME; 20 mg/kg) produced an over 50% inhibition of the dose-related salivation induced by metacholine. Similar results were produced with L-NG-monomethyl-L-arginine (L-NMMA; 5 mg/kg). The salivation induced by much lower molar doses of substance P was dramatically greater than that obtained with metacholine. The response to substance P was almost completely inhibited by L-NMMA at the lowest dose (0.3 mg/kg), but at higher doses (1 mg/kg), the inhibition was only around 60% and at the highest dose (3 mg/kg) only about 20%. In control rats, there were roughly equal amounts of calcium-dependent and calcium-independent NOS in the gland at this time. At the end of the experiment, the effect of the inhibitor of NOS, L-NMMA, on the NOS activity in the submandibular gland was determined. At this time, the Ca2+-dependent NOS was decreased and the Ca2+-independent NO was increased. The prior injection of L-NMMA reduced calcium-dependent NOS activity by approximately 70% but calcium-independent activity by only 30%. These results indicate that, at least at the end of the experiment, the blockade of NOS imposed by NMMA was incomplete. This could account in part for the failure of the inhibitors to block completely the stimulatory effect of the two secretagogues. Analysis of the distribution of NOS in the salivary gland revealed that it was not present in the acinar cells, but in neural terminals within the gland and also in the ductile system which contained neural (n) NOS in the apical membrane of the excretory and striated ducts, the cytoplasm of granular convoluted tubules and, to a lesser extent, in the cytoplasm of excretory and striated ducts. Macrophage (inducible) NOS was also found not only in the macrophages, but also in the tubules and ducts. Since drugs were used that would act on the receptors in the gland, the role of NO in our conditions is probably mediated by nNOS and iNOS in the ductile and tubular structures. Since iNOS would already be active, it is unlikely to play a role in this acute secretory activity. Rather the nNOS in these non-neural cells is probably activated by muscarinic or K1 receptors by metacholine and substance P, respectively, leading to an increase in intracellular free calcium that activates NOS leading to the generation of cGMP that opens ion channels to initiate the secretory process.