A single large po or sc dose of indomethacin, 80 times the antiinflammatory dose, produced a high incidence of intestinal perforations 72 hr after administration to fed rats; the perforations were confined to the mid-portion of the small intestine. When food was withheld from indomethacin-treated rats, the lesion incidence after a dose of 16 mg/kg po was reduced from 100% to zero. In a 72-hr study, it was found that food deprivation only on the day of drug administration prevented the intestinal perforations, while feeding the rats on the day of drug administration produced 100% ulcer incidence. Lesions produced by sc drug administration could also be prevented by starvation. The incidence of intestinal perforation was altered by the duration of food deprivation and the amount of food consumed. Intestinal ulcers could also be prevented, after either po or sc administration, by ligation of the bile duct. It appeared possible that there was a relationship between food intake, bile flow, and intestinal ulcers. Bile duct ligation reduced the incidence of intestinal perforation after indomethacin from 100% to zero. A time response curve indicated that bile duct ligation up to 8 hours after drug administration reduced the incidence of perforation; however, a 60% incidence of intestinal lesions was found when the bile duct was ligated 3 hours after indomethacin was given. When segments of the small intestine were isolated as Thiry loops, ulceration was prevented in the loops but occurred in the anastomosed intestine.