ABSTRACT. To evaluate the concept that changes in colonic blood flow will predictably alter the absorption of colonic gas, we measured the pulmonary clearance rate of helium (CHe) which was instilled rectally into the colon of rabbits at a dose of 2 ml/kg. CHE reached a plateau after 20 min at 109 nmol/min/kg. The underlying principle for the technique is that pulmonary clearance of an inert gas instilled into the colon is predictably altered by changes in colonic blood flow and that the state of colonic blood flow in the segment measured is at least an indicator of regional gut perfusion.The concept of using bowel gas absorption to reflect bowel perfusion has previously been applied in both animals and in man. Animal studies examining the relationship between intestinal gas absorption and bowel blood flow have typically used techniques requiring surgical exteriorization of a bowel segment (1 -5). Similar bowel exteriorization studies have been performed on adult humans undergoing laparotomy and have involved measuring pulmonary gas excretion following gas instillation into the jejunum or colon (6). In the present study, He, a passively absorbed inert gas, was instilled rectally into the colon of rabbits, and its pulmonary clearance was measured by end-expiratory breath sampling.We then examined changes in C H~ during hypoxemia, and insult which is known to cause decreased mesenteric perfusion and bowel ischemia (7). This relationship between pulmonary clearance of colonically absorbed gas and effective subvillus colonic blood flow is discussed in terms of a currently available model for inert gas absorption, and it can be shown that endexpiratory levels should reflect changes in subvillus blood flow in minute-ventilation is known.The capability to measure clinically intestinal blood flow has many potential benefits in patients who have had known hypoxemic or hypotensive insults. One such group of patients would be small premature infants whose adequate nutritional status depends largely on the early institution of enteral feedings. Because prenatal and natal physiologic events may create states of intestinal ischemia which can predispose to bowel injury, development of a method for monitoring intestinal hypoperfusion may help reduce morbidity by giving useful clinical information to direct neonatal bowel resuscitation and to establish the timeliness of gastrointestinal use. It is possible that the early use of the hypoperfused gut for enteral nutrition may result in the specific complication of necrotizing enterocolitis, the etiology of which is believed to be a combination of bowel ischemia, luminal carbohydrate substrate, and bacterial growth with fermentation and tissue invasion. Our study was therefore undertaken to develop a method of assessing changes in intestinal circulation