Aggressive treatment with H(2) receptor blocking agents and/or antacids has been advocated as effective prophylaxis against and treatment for "stress ulcer," based on the logical but infrequently tested assumption that the severity of the disease is critically determined by the concentration of intraluminal acid. The present study investigated this assumption in a model which employed topical acid, topical bile acid and mucosal ischemia to induce ulcerogenesis. With vascularized, chambered ex vivo wedges of canine proximal gastric wall, groups of animals were studied during three sequential periods using topical test solutions (TS) containing either 0 mM, 100 mM or 160 mM HCI. During period 1, mucosae were exposed to TS alone; during period 2, either to TS containing 1 mM sodium taurocholate (TC) or to TS and concomitant vasopressin infusion (VP); and during period 3, to TS + TC + VP. Parameters evaluated included net H(+) flux ( big up tri, openH(+)), aminopyrine clearance (AC), a measure of mucosal blood flow, net TC flux ( big up tri, openTC) and the lesion index, graded 0-5. The data indicate that in nonischemic mucosa exposed to constant [TC], AC was significantly increased, big up tri, openH(+) ("back-diffusion") increased as a linear function of [H(+)] and no lesions were observed. Under the same circumstances in ischemic mucosa, big up tri, openH(+) increased as linear function of [H(+)]. As a consequence, lesion severity was also a linear function of [H(+)]. big up tri, openTC was enhanced at low pH but bore no relation to the degree of mucosal damage induced. Assuming applicability of the model, these studies provide support for the use of H(2) receptor blocking agents and/or antacids to prevent or ameliorate "stress ulcer" disease.