During uncontrolled hemorrhagic shock (UHS) in acute animals models, attempts to achieve normotension with i.v. fluid resuscitation (FR) caused further bleeding and higher acute mortality. In the absence of a published clinically realistic long-term animal outcome study of UHS, we developed such a model in rats. In the preliminary study, phase I of the model involved 60 min of simulated 'pre-hospital' UHS by tail amputation and different FR regimens. Phase II involved 120 min of simulated 'hospital' treatment with hemostasis and all-out FR, including blood infusion. Phase III involved observing recovery and survival to 72 h (3 days). Rats were maintained under very light N2O-O2-halothane anesthesia and spontaneous breathing via mask during phases I and II and were awake during phase III. Tail amputation-induced UHS alone, studied in 4 groups of 10 rats each, resulted in unpredictable spontaneous hemostasis and great variability in shed blood volume, severity of shock, and mortality. The final model, which achieved consistent blood loss and outcome, included an initial volume-controlled hemorrhage of 3 ml/100 g over 15 min and untreated HS for another 15 min, followed by tail amputation for UHS over another 60 min. This phase I of 90 min was followed by phase II of 60 min. In group 1, without FR in phases I and II, all 10 rats died by 12 h. In group 2, without FR in phase I and hemostasis plus all-out FR with lactated Ringer's solution and blood to hematocrit (Hct) 30% in phase II, 5 of 10 rats died at the end of phase I and 9 of 10 died at the end of phase III. This final volume-initiated UHS model may be suitable for comparing different pre-hospital treatment modalities in terms of outcome.