Allogeneic blood resuscitation is the treatment of choice for hemorrhagic shock. When blood is unavailable, plasma expanders, including crystalloids, colloids, and blood substitutes, may be used. Another treatment modality is vasopressin, a vasoconstrictor administered to redistribute blood flow, increase venous return, and maintain adequate cardiac output. While much information exists on systemic function and oxygenation characteristics following treatment with these resuscitants, data on their effects on the microcirculation and correlation of real-time microvascular changes with changes in systemic function and oxygenation in the same animal are lacking. In this study, real-time microvascular changes during hemorrhagic shock treatment were correlated with systemic function and oxygenation changes in a canine hemorrhagic shock model (50-55% total blood loss with a MAP of 45-50 mmHg as a clinical criterion). Following splenectomy and hemorrhage, the dogs were assigned to five resuscitation groups: autologous/shed blood, hemoglobin-based oxygen carrier/Oxyglobin, crystalloid/saline, colloid/Hespan (6% hetastarch), and vasopressin. Systemic function and oxygenation changes were continuously monitored and periodically measured (during various phases of the study) using standard operating room protocols. Computer-assisted intravital video-microscopy was used to objectively analyze and quantify real-time microvascular changes (diameter, red-cell velocity) in the conjunctival microcirculation. Measurements were made during pre-hemorrhagic (baseline), post-hemorrhagic (pre-resuscitation), and post-resuscitation phases of the study. Pre-hemorrhagic microvascular variables were similar in all dogs (venular diameter = 42+/-4 microm, red-cell velocity = 0.55+/-0.5 mm/sec). All dogs showed significant (P < 0.05) post-hemorrhagic microvascular changes: approximately 20% decrease in venular diameter and approximately 30% increase in red-cell velocity, indicative of sympathetic effects arising from substantial blood loss. Microvascular changes correlated with post-hemorrhagic systemic function and oxygenation changes. All resuscitation modalities except vasopressin restored microvascular and systemic function changes close to pre-hemorrhagic values. However, only autologous blood restored oxygenation changes to pre-hemorrhagic levels. Vasopressin treatment resulted in further decreases in venular diameter (approximately 50%) as well as red-cell velocity (approximately 70%) without improving cardiac output. Our results suggested that volume replenishment - not oxygen-carrying capability - played an important role in pre-hospital/en route treatment for hemorrhagic shock. Vasopressin treatment resulted in inadvertent detrimental outcome without the intended benefit.