Application of an irritant to the surface of the gastric mucosa confers protection against subsequent application of a damaging agent ("adaptive cytoprotection"). The possibility that this adaptive response is mediated via the release of nonprostaglandin mediators was examined using an ex vivo gastric chamber model, in which the irritant (1 M sodium chloride) could be applied to only one-half of the exposed mucosa. Rats were pretreated with indomethacin and one of various receptor antagonists, including cimetidine, pyrilamine, atropine, propranolol, phentolamine, cyproheptadine, haloperidol, and BN52021. In control studies, application of hypertonic saline to the left side of the mucosa reduced hemorrhagic damage induced by subsequent application of 70% ethanol by 71% +/- 4% (p less than 0.01). Of the receptor antagonists tested, only cyproheptadine and haloperidol significantly attenuated the degree of protection afforded by exposure to hypertonic saline. Subsequent dose-response studies with other serotonergic and dopaminergic antagonists suggested that dopaminergic receptors are involved in the adaptive response to the irritant. Topical application of a dopamine agonist, bromocriptine, or L-beta-3,4-dihydroxyphenylalanine, significantly reduced the extent of damage induced by subsequent application of 70% ethanol. The results of the gastric chamber studies were confirmed in conscious rats in which the irritant, dopamine agonist, and ethanol were administered orally. These results therefore suggest a role for endogenous dopamine as a mediator of the adaptive cytoprotection phenomenon.