SUMMARY To determine whether hypoxia directly affects pulmonary microvascular filtration of fluid or permeability to plasma proteins, we measured steady state lung lymph flow and protein transport in eight unanesthetized sheep breathing 10% O 2 in N» for 4 hours. We also studied three sheep breathing the same gas mixture for 48 hours. We surgically prepared the sheep to isolate and collect lung lymph and to measure average pulmonary arterial (P pa ) and left atrial (P la ) pressures. We placed a balloon catheter in the left atrium to elevate P la . After recovery, the sheep breathed air through a tracheostomy for 2-4 hours, followed by 4 or 48 hours of hypoxia. In 13 4-hour studies, the average arterial Po 2 fell from 97 to 38 torr; P pa rose from 20 to 33 cm H,O; and lung lymph flow and lymph protein flow were unchanged. We also found that during 48-hour hypoxia, with a sustained elevation in P pa and a decline in P la , lymph flow and protein flow did not increase. In four sheep, we also raised Pi a for 4 hours, followed by 4 hours of hypoxia with elevated P la . Again, despite the added stress of elevated P la , we found that lymph flow and lymph protein flow remained constant during hypoxia. We conclude that severe alveolar hypoxia, for 4 or 48 hours, alone or with increased pulmonary microvascular pressure, produced no change in lung fluid filtration or protein permeability, a finding supported by normal postmortem histology and extravascular lung water content.IN 1945, DRINKER concluded that "oxygen lack is the most potent and elusive cause of abnormal leakage from the lung capillaries," 1 based on the observations that flow from the right lymphatic duct of anesthetized dogs increased when the dogs breathed a 10% oxygen gas mixture.2 -3 Courtice and Korner 4 reported in 1952 that hypoxia predisposed rabbits to pulmonary edema induced by infusions of Ringer's solution, but the authors attributed the edema to heart failure and not altered vascular permeability to plasma proteins.Other investigators have been unable to substantiate Drinker's conclusion. Haddy et al. 5 showed in anesthetized dogs that hypoxia alone did not produce pulmonary edema, but only did so in the presence of elevated pulmonary venous pressure. Hemingway 6 exposed guinea pigs to gas mixtures containing as little as 2% oxygen, but found no evidence of pulmonary edema when he killed the guinea pigs and examined their lungs.Using isolated, perfused dog lungs, Nicoloff et al. 7 were unable to demonstrate an increase in lung weight with extreme hypoxia. Goodale et al. 8 found that even total absence of oxygen in the inspired gas did not alter the permeability of the alveolocapillary membrane to tracer albumin in isolated, perfused dog lungs. Fisher et al.9 saw no changes in the ultrastructure of the alveolar septum of intact dog lungs ventilated with nitrogen for 3-7 hours, and Teplitz et al. 10 were unable to elicit pulmonary edema in rats with hypobaric hypoxia equivalent to 7% oxygen in the inspired gas for up to 30 hours.Despite all of th...