OBJECTIVE The present study was undertaken to assess the fluid balance and hemodynamic effects during the first 8 hrs of resuscitation in animals with a large body surface area burn, using lactated Ringer's solution, 6% hetastarch, and hypertonic saline dextran. METHODS A prospective, blinded, controlled, terminal study, using anesthetized animals. The initial baseline period was followed by scald injury, and three different treatment regimens were administered from coded bags to achieve a physiologic end point. METHODS University laboratory. METHODS Eighteen female sheep (35 to 45 kg) were anesthetized with isoflurane. METHODS Test solutions (10 mL/kg of either lactated Ringer's solution, hetastarch, or hypertonic saline dextran) were infused 30 mins after the scald injury at a rate to restore and maintain the baseline oxygen delivery (DO2) value. RESULTS Cardiovascular hemodynamics, plasma sodium concentration, plasma colloid osmotic pressure, and fluid balance were measured before and after scalding and resuscitation. After the initial 10-mL/kg test solution dose was given, lactated Ringer's solution was infused to achieve the same end point of baseline DO2 for the remainder of the 8 hrs. The scald caused an initial 30% reduction in cardiac output, a 20% reduction in mean arterial pressure, and 10% to 15% increase in hematocrit. All three test solutions restored and maintained baseline DO2 within 1 hr. However, hetastarch and hypertonic saline dextran reduced the net fluid volume over 8 hrs by 48% and 74%, respectively, compared with lactated Ringer's solution. Edema in the burn wound was not affected by treatment, while hypertonic saline dextran reduced edema in nonburned skin compared with both lactated Ringer's solution and hypertonic saline dextran. Plasma colloid osmotic pressure was significantly higher in the hetastarch and hypertonic saline dextran groups. A continuous decrease in plasma sodium concentrations from baseline values (140 to 145 mmol/L) was measured in the lactated Ringer's solution and hetastarch groups (130 to 133 mmol/L) over 8 hrs. Plasma sodium concentrations in the group receiving hypertonic saline dextran were increased (150 to 155 mmol/L) at 4 hrs, but returned toward baseline by 8 hrs. CONCLUSIONS Net volume loading can be reduced markedly by initial resuscitation of large body surface area burn injury using a colloid (hetastarch), and can be further reduced by use of hypertonic saline colloid. Hyponatremia was apparent in the isotonic crystalloid- and colloid-treated animals, but not in those animals treated with hypertonic saline colloid.