Fifty-seven dogs were subjected to hemorrhagic hypotension by a variety of protocols. Histologic pulmonary changes were studied using the light microscope. Of these 57 dogs, 21 had no demonstrable lesions, 8 had minimal changes, and 28 had moderate or severe lesions, all of a focal nature. No correlation was found between the presence of lesions and mean systemic arterial pressure during shock, the udration of the hemorrhagic period, the fate of the animal, preoperative hematocrits and blood volumes, mean postreinfusion arterial pressure, whether the animals were mongrels or purebred beagles, whether they were awake or sedated, whether they breathed spontaneously or were artifically ventilated, whether they had undergone previous splenectomy or not, whether hilar stripping was performed or not, and finally, whether blood was reinfused after hemorrhage or not. Thus we conclude that multiple factors may exert a harmful effect on the lung in hemorrhagic shock, and that shock probably makes the lungs more vulnerable to other injurious agents rather than there being one single pathogenetic mechanism for the pulmonary damage. The term "adult respiratory distress syndrome" rather than "shock lung" is best used for the human clinical entity since it implies a complex ettiology rather than a discrete pulmonary lesion produced by a single pathogenetic mechanism.