W e studied anesthetized sheep to determine the relationship between increased permeability pulmonary edema and the development and mechanism of pleural effusion formation. In 12 sheep with intact, closed thoraces, we studied the time course of pleural liquid formation after 0.12 ml/kg i.v. oleic acid. After 1 h, there were no pleural effusions, even though extravascular lung water increased 50% to 6.0 +/- 0.7 g/g dry lung. By 3 h pleural effusions had formed, they reached a maximum at 5 h (48.5 +/- 16.9 ml/thorax), and at 8 h there was no additional accumulation of pleural liquid (45.5 +/- 16.9 ml). Morphologic studies by light and electron microscopy demonstrated subpleural edema but no detectable injury to the visceral pleura, suggesting that the pleural liquid originated from the lung and not the pleura. In nine sheep, we quantified the rate of formation of pleural liquid by enclosing one lung in a plastic bag. By comparing in the same sheep the volume of pleural liquid collected from the enclosed lung to the volume found in the opposite intact chest, we estimated the rate of liquid absorption from the intact chest to be 0.32 ml/(kg.h); we had previously reported a liquid absorption rate of 0.28 ml/(kg.h) in normal sheep. These studies also supported the conclusion that the majority of the pleural liquid originated from the lung because we could account for all of the pleural liquid that was formed and cleared. The volume of pleural liquid collected from the enclosed lungs was equal to 21% of the excess lung liquid that formed after oleic acid-induced lung injury. Thus, the pleural space and parietal pleural lymphatic pathways are important pathways for the clearance of pulmonary edema liquid after experimentally induced increased permeability pulmonary edema.