In rabbits, laser Doppler flow probes were placed in the jejunum and on the renal cortex. Pulsed Doppler probes were implanted on the abdominal aorta and superior mesenteric and femoral arteries for measuring blood flow velocity. Cardiac output was measured by thermal dilution. Either 30% or 40% of the calculated blood volume was withdrawn through a carotid catheter. After 30 or 60 minutes, an initial bolus of either lactated Ringer's (LR, 16 ml/kg) or 7.5% hypertonic saline/6% dextran 70 (HSD; 4 ml/kg) IV was followed by unlimited IV LR (administered as rapidly as possible) to restore systemic arterial blood pressure to the prehemorrhage levels. With HSD, arterial pressure corrected more rapidly (p less than 0.05), and the initial hemodilution was greater (p less than 0.05), but there were no differences by two hours. With HSD, cardiac output (90%-100% vs. 130%-160% of control; p less than 0.05), plasma Na+ (139-140 mM vs. 146-148 mM; p less than 0.05) and plasma osmolarity (292-295 mOsm vs. 308-310 mOsm; p less than 0.05) were all significantly higher than the values with LR, but there was no effect on blood flow velocities through the infrarenal aorta, femoral artery, or superior mesenteric artery. Renal cortical perfusion (56% vs. 97% of control; p less than 0.05) and jejunal mucosal perfusion (83% vs. 162% of control; p less than 0.05) were significantly higher with HSD. HSD had no detectable effect on bacterial translocation at 24 hours. Thus: 1) HSD restores blood flow more rapidly to the gut mucosal and kidney microcirculations than initial resuscitation with LR; 2) the mechanism could be associated with a transient hemodilution and persistent increases in plasma Na and osmolarity, which reduce hemorrhage-induced cell swelling and blood viscosity changes; and 3) laser Doppler analysis could aid in the diagnosis of reperfusion injury after shock.