We evaluated an endovascular cooling method to modulate core temperature in trauma swine models with and without fluid support. Anesthetized swine (N = 80) were uninjured (SHAM) or injured through a bone fracture plus soft tissue injury or an uncontrolled hemorrhage and then subdivided to target body temperatures of 38°C (normothermia) or 33°C (hypothermia) by using a Thermogard endovascular cooling device (Zoll Medical). Temperature regulation began simultaneously at onset of injury (T0). Body temperatures were recorded from a rectal probe (Rec Temp) and from a central pulmonary artery catheter (PA Temp). At T15, swine received 500 mL IV Hextend over 30 minutes or no treatment (NONE) with continued monitoring until 3 hours from injury. Hypothermia was attained in 105 ± 39 minutes, at a cooling rate of -0.061°C ± 0.007°C/min for NONE injury groups. Postinjury Hextend administration resulted in faster cooling (-0.080°C ± 0.006°C/min); target temperature was reached in 83 ± 11 minutes (p < 0.05). During active cooling, body temperature measured by the PA Temp was significantly cooler than the Rec Temp due to the probe's closer proximity to the blood-cooling catheter balloons (p < 0.05). This difference was smaller in SHAM and fluid-supported injury groups (1.1°C ± 0.4°C) versus injured NONE groups (2.1°C ± 0.3°C). Target temperatures were correctly maintained thereafter in all groups. In normothermia groups, there was a small initial transient overshoot to maintain 38°C. Despite the noticeable difference between PA Temp and Rec Temp until target temperature was attained, this endovascular method can safely induce moderate hypothermia in anesthetized swine. However, likely due to their compromised hemodynamic state, cooling in hypovolemic and/or injured patients will be different from those without injury or those that also received fluids.