A B S T R A C T Radiolabeled, enzymatically active or chloromethyl ketone-inactivated porcine pancreatic elastase was endotracheally instilled into hamsters. Gel filtration of the bronchopulmonary lavage fluid revealed two major radioactive fractions: one, eluting at 780,000 daltons, corresponding to an a-macroglobulinpancreatic elastase complex, and another, at 68,000 daltons, corresponding to an a-1-protease inhibitorpancreatic elastase complex. Elastolytic activity was recovered in the bronchopulmonary lavage fluid up to 4 d after elastase instillation and was associated with the a-macroglobulin-pancreatic elastase complex. Small amounts of this complex were recovered 15 d after instillation. When <1% (1.5-1.7 ,ug) of the usual dose of elastase was instilled into hamsters, the major radioactive complex was a-1-protease inhibitor-pancreatic elastase complex, and little or no elastolytic activity was found in the lavage fluid. In contrast to the instillation of 220 tsg of elastase, no disease or hemorrhagic reaction was detected with this low dose, and without hemorrhage only insignificant amounts of amacroglobulin-pancreatic elastase complexes were recovered from the lungs. To study the interaction of circulating antiproteases with elastase, hamster plasma was allowed to interact directly with the radiolabeled elastase; a-macroglobulin bound much more of the elastase than a-1-protease inhibitor, confirming the findings in the lung lavage experiments. The hamster's susceptibility to pancreatic elastase-induced emphysema may depend on the preferential binding of elastase to a-macroglobulin, which protects the elastolytic potential, rather than to a-1-protease inhibitor, which