Since the demonstration of regional perfusion for cancer chemotherapy was firsts hown by Oscar Creech (in 1958) with an extracorporeal circuit, regional perfusion has been attempted by manyclinicians. After these perfusions were begun with anti-cancerous agents, it has been applied to thebrain, lung, extremities, pelvic organs and mid-gut. The present paper is designed to study the tissue damage caused by Thio-TEPA and Nitrogen Mustard-N-oxid (N. M. O.) extracorporeal perfusions. In 1960, Shingleton demonstrated a method of abdominal perfusion via. the abdominal aorta, butthe author designed a new method via, the gastroepiploic arteries, and gained satisfactory perfusionwith adequate conditions in dogs. As a result, various tissue damage was observed in the cases where over 2.5 mg./kg. of Thio TEPAwere given in the perfusions. The types of damage were edema, hemorrhage associated with erosion of the mucous membrane, necrosis, necrobiosis and fibrosis of the mucous layer. Some cases in which 50 mg./kg. of ThioTEPA were given, died of perforation of the stomach. However, no remarkable tissue damage could beseen in the cases where less than 2.0mg./kg. were given. In the case of N. M. O., tissue damage was observed in the cases where over 5.0 mg./kg. were given. In these cases, edema, hemorrhage. and necrosis could be seen in ten perfused area. Some casesdied of perforation of the stomach as was found with Thio-TEPA. However, no remarkable tissuedamage could be seen in the cases of where less than 2.5 mg./kg. of N. M. O. were given. There is evidence that Thio TEPA and N. M. O. cause hemorrhage of the capillary arteries whichdevelop necrobiosis or necrosis of tissues, and some are turned to fibrosis. These tissue damages are chiefly made by three factors: the first is toxic reaction of the agents: the second is perfusion itself which causes anoxia of tissues, and the third is the degree of recoveryof blood circulation after the perfusions. At the end of this paper the author concluded that, at present this regional perfusion of the stomachis not recommended to be utilized for clinical cases as cancer chemotherapy because its perfused areais too limited to reach the spreading stomach cancer metastasis and because of the easily encounteredtissue damages.