This study was undertaken in an attempt to establish a method of bleeding rats and autotransfusing the blood removed from them which would be associated with a mortality of near 50%. The use of different combinations of fluids as therapy for hemorrhage in the rat could be assessed optimally, since the likelihood of a given method of treatment being both better or worse than autotransfusing all the blood removed can best be determined when the latter is associated with a mortality of 50%. A previous study from this laboratory reported a hemorrhagic shock model in the rat which was based upon bleeding to a mean blood pressure of 30 mm Hg for 60 minutes.1 This model was associated with a mortality of only 24% among rats surviving to be treated by returning all the blood removed from them. When the duration of hypotension was lengthened more rats died before