T he present study is an attempt to determine (a) if the lung actively secretes protein into the surface-active fraction of lung lavage returns; (b) if there are interspecies differences in this secretory activity; and (c) if the amount of nonradioactive protein in the lavage surface-active fraction shows interspecies variation. I found that pilocarpine stimulates the release of radioactive protein into the lavage surface-active fraction of rabbits and that this pilocarpine effect is completely blocked by atropine. Inhibition of lung oxygen consumption by iodoacetate is associaged with a dose-dependent inhibition of the pilocarpine-induced secretion. Microtubules may be involved in this secretory process because colchicine inhibits the pilocarpine effect. Of the radioactive protein in the total surface-active fraction (tissue plus lavage returns), a greater percent appears in the lavage surface-active fraction at 2 and 4 h, after a pulsed injection [U-14C] leucine, in the mouse than in the rat, which in turn has a greater amount than the rabbit. There is also a difference in the amount of nonradioactive protein per square meter of alveolar surface area in the lavage surface-active fraction of different species: mouse greater than rabbit greater than cat greater than dog. The amount of nonradioactive protein per square meter of alveolar surface area in the lavage surface-active fraction is directly proportional to the species respiratory rate; the log of the nonradioactive protein in the lavage surface-active fraction is inversely proportional to the log of the species alveolar diameter. I conclude that the lung actively secretes protein into the lavage surface-active fraction, that this secretion is under neurohumoral regulation, and that respiratory rate and alveolar size may influence this secretory activity and the amount of protein in this surface-active fraction.