Background: Microdialysis was used to study how ischemia-evoked gastric mucosal injury affects rat stomach histamine, which resides in ECL cells (~80%) and mast cells (~20%). Materials and methods: A microdialysis probe was inserted into the gastric submucosa (pentobarbital anesthesia) and the celiac artery was clamped (ischemia, 30 min), followed by removal of the clamp (reperfusion, 60 min). Microdialysate samples were collected every 10 min, and histamine was determined by enzyme-linked immunosorbent assay. In addition, we studied the long-term effects of ischemia on the oxyntic mucosal histamine concentration and histidine decarboxylase activity in omeprazole-treated rats. Results: The lesioned area in the mucosa was measured after 30 min of arterial clamping, and after 30 or 60 min of reperfusion. Lesions induced by the ischemia were enlarged upon removal of the clamp. The microdialysate histamine concentration increased immediately upon clamping (50-fold rise within 30 min). After removing the clamp, the microdialysate histamine concentration declined promptly. Mast-cell deficient rats responded to ischemia-reperfusion much like wild type rats with respect to both mucosal lesions and histamine mobilization. Subcutaneous infusion (4 days) of the irreversible inhibitor of histidine decarboxylase, -fluoromethylhistidine, which is known to eliminate histamine from ECL cells (but not from mast cells), prevented the rise in microdialysate histamine and greatly reduced the area of mucosal lesions. A single subcutaneous injection of the histamine H₂ receptor antagonist cimetidine or per oral administration of the proton pump inhibitor omeprazole for 4 days prevented the mucosal lesions but not the mobilization of histamine. The histidine decarboxylase activity of the ECL cells was raised by the omeprazole treatment (because of the hypergastrinemia). The enzyme activity in these rats was lowered by the ischemia and returned to pre-ischemic values 9 days later (continued omeprazole treatment). The oxyntic mucosal histamine concentration was low shortly after the ischemia and then returned to normal values. Conclusions: Ischemia of the celiac artery mobilized large amounts of histamine. We propose that the mobilized histamine comes from the ECL cells, and that mast cells do not contribute to either the lesions or the histamine response. The low histamine response to ischemia in rats pretreated with -fluoromethylhistidine supports this view. Pharmacological blockade of acid secretion (cimetidine or omeprazole) prevented the lesions induced by ischemia-reperfusion insult but not the mobilization of histamine. We suggest that the lesions develop not because of mobilization of histamine per se but because of ischemia + reperfusion + gastric acid per se. Mobilization of ECL-cell histamine is a consequence of the ischemia but occurs independently of the gross mucosal lesions. The prompt reduction of the oxyntic mucosal histidine decarboxylase activity (omeprazole-treated rats) in response to 30 min of ischemia probably reflects ECL-cell damage. The return of the enzyme activity 9 days later suggests that the cells recover from the damage.