Trauma registers show that hypothermia (HT) is an independent risk factor for death during hemorrhagic shock, although experimental animal studies indicate that HT may be beneficial during these conditions. However, the animal models were not designed to detect the expected increase in bleeding caused by HT. In a new model for uncontrolled bleeding, 40 Sprague-Dawley rats were exposed to a standardized femoral artery injury and randomized to either normothermia or HT. Ketamine/midazolam was used to minimize hemodynamic changes due to the anesthesia. The hypothermic rats were cooled to 30°C and rewarmed again at 90 min. The study period was 3 h. The incidence, onset time, duration, and volume of bleedings as well as hemodynamic and metabolic changes were recorded. There was no difference between groups with respect to the initial bleeding. Rebleedings occurred among 60% of the animals in both groups. Hypothermic rebleeders had more, larger, and longer rebleedings, resulting in a total rebleeding volume amounting to 41% of their estimated blood volume. The corresponding figure for the normothermic rebleeders was 3% (P < 0.001). Total rebleeding volume was significantly larger in the hypothermic group, even at body temperatures greater than 35°C. We conclude that the risk of rebleeding from a femoral injury is greater in the presence of cooling and HT. The larger rebleeding volumes seen even at body temperatures greater than 35°C indicate that factors other than temperature-induced coagulopathy also contributed to the increased hemorrhage.