Objective Current hemorrhagic animal models often have a wide range of blood pressure fluctuation and limited brain tissue damage. We aimed to develop a prolonged severe hemorrhagic shock model without significant blood pressure fluctuation in rats. Methods A 10 mL syringe was fixed on a wall at 47. 58 cm above a surgical operating board to serve as a blood reservoir. A hemorrhagic shock model was induced by automatically withdrawing blood into the 10 mL syringe. Mean arterial blood pressure (MAP) was controlled by adjusting the blood level inside the syringe. Twenty-eight rats were randomly divided into four groups in accordance with different resuscitations. The hemorrhagic shock episode lasted for 3 h. The physiological conditions and hemodynamic indexes of rats in each group were monitored in real-time and arterial blood gas was analyzed before and after the shock. Brain hippocampal neuronal damage was assessed by HE staining. Results MAP was accurately maintained at 31~ 35 mm Hg by keeping the blood level inside the syringe at 47. 58 cm above the surgical operating board throughout the 3 h hemorrhagic shock episode. The rats in all groups had survived at the end of the 3 h shock. Rats in the non-resuscitation group had died within 2 h after the end of the shock and those in resuscitation (RB and RR) groups had survived within 3 h after resuscitation. There was a significant decrease in blood pH and increase in blood lactate. Maximum blood loss had occurred at 20 ~ 30 min after hemorrhagic shock and the maximum blood loss was very close to the final blood loss. Pathology revealed marked hippocampal neuronal damage in all rats. Conclusions The syringe blood reservoir is useful to establish lengthy severe hemorrhagic shock with accurate control of MAP. This highly reproducible hemorrhagic shock model induces reliable brain damage to investigate pathophysiological changes and therapeutic effects of hemorrhagic shock.