Previous work in our laboratory has demonstrated that HSD is an effective small-volume resuscitation fluid for the treatment of hemorrhagic hypotension, but limitations to its usefulness in severe hemorrhage have not been explored. In the present study, animals (N = 12) were bled from an arterial line at a rate of 1 mL/kg/min until continuously monitored aortic blood flow was reduced to one-half its baseline value, and then they were immediately resuscitated with 7.5% NaCl/6% dextran 70 (hypertonic saline dextran, 4 mL/kg) administered intravenously over 3 min. After recording the maximum improvement in blood pressure, blood samples were obtained and the hemorrhage-resuscitation sequence was repeated until no further measurable increase in cardiac index or blood pressure could be elicited by resuscitation. In the majority of the animals, cardiac index and right and left ventricular stroke work could be improved at least through two bleedings and resuscitation. These improvements sufficed to increase oxygen delivery and consumption, despite the decreases in hematocrit induced by bleeding, transcapillary refill, and asanguinous fluid administration. Under these severe hemorrhage conditions, the acid-base imbalance was not improved by hypertonic saline dextran, and the rate of increase in acidosis was not affected by its administration. We observed a progressive decrease in base excess from +1.35+/-3.19 (mean +/- standard error) to -12.9+/-2.1 mEq/L even when resuscitation improved oxygen consumption significantly by 95+/-20%. In animals that survived as many as three bleedings and resuscitation, the depletion of buffering capacity of the blood was most predominant, and bicarbonate reached a nadir of 7.62 mEq/L with a base excess of -22.4 mEq/L. It is evident that restoration of perfusion in shock treats only a portion of the physiologic dysfunction, leaving major metabolic derangements uncorrected.