Introduction: In recent years we have developed and used UHS models via tail cut to simulate clinical conditions. A 3 phase outcome model in rats with UHS 75min or 90min resulted in a high mortality with great variability. The aim of this study was to modify our model to achieve, after UHS 90 min at normothermia, survival to 72h in about 50% of rats, without systemic heparin, and with use of donor blood for minimal fluid resuscitation (FR) to MAP 40 mmHg. Methods: Under light halothane anesthesia (N2:O2, 75%:25%), UHS phase I was induced by a pump driven blood withdrawal of 3ml/100g over 15 min, tail amputation, and FR with donor blood from 20 min to 90 min, keeping MAP at ≥40 mmHg. Phase II included hemostasis and FR to normotension with shed blood and monitoring for 3h. After awakening, observation was to 72h (phase III). Pilot studies revealed necessary modifications in phase I (topical lidocaine/heparin on tail stump, more rapid start of FR, pump controlled infusion), and in phase II (speed of shed blood infusion, correction of hypoglycemia, titrated normal saline infusion). Hemodynamic parameters, blood loss, blood gases, lactate, glucose, liver dysoxia (liver surface pCO 2) were monitored. Necropsy was performed at early death or euthanasia at 72h. Results: The survival rates were 10/10 at the end of phase I, 8/10 in phase II and 4/10 in phase III. The total blood loss was 4.5±0.4ml/100g. In phase I, 2.2±0.8ml/100g donor blood had to be given to keep MAP at ≥40 mmHg. Liver pCO2 increased rapidly from 66±15 to 223±73 mmHg at 20 min, decreased to 122±67 mm Hg at end of phase I and reached baseline at end of phase II. Additional glucose was needed to achieve normoglycemia and survival of phase II. Necropsies revealed gut necrosis in non-survivors. Conclusions: In a severe UHS 90min outcome model, modifications of our previously almost 100% lethal model achieved 4/10 long-term survival. This model appears suitable to evaluate beneficial or deleterious outcome effects of different resuscitation fluids and medications to be evaluated for phases I and II.