Objective The objective of this study was to test the therapeutic effect of carbon monoxide polyhemoglobin (polyCOHb) in haemorrhagic shock/resuscitation and its underlying mechanisms. Methods 48 rats were divided into two experimental parts, and 36 rats in the first experiment and 12 rats in the second experiment. In the first experimental part, 36 animals were randomly assigned to the following groups: hydroxyethyl starch group (HES group, n = 12), polyhemoglobin group (polyHb group, n = 12), and carbon monoxide polyhemoglobin group (polyCOHb group, n = 12). In the second experimental part, 12 animals were randomly assigned to the following groups: polyHb group (n = 6), and polyCOHb group (n = 6). Then the anaesthetised rats were haemorrhaged by withdrawing 50% of the animal’s blood volume (BV), and resuscitated to the same volume of the animal’s withdrawing BV with HES, polyHb, polyCOHb. In the first experimental part, the 72h survival rates of each groups animals were measured. In the second experimental part, the rats’ mean arterial pressure (MAP), heart rate (HR), blood gas levels and other indicators were dynamically monitored in baseline, haemorrhagic shock (HS), at 0point resuscitation (RS 0h) and after 1 h resuscitation (RS 1h). The concentrations of malondialdehyde (MDA), superoxide dismutase (SOD), tumour necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) were measured by ELISA kits in both groups of rats at RS 1h. Changes in pathological sections were examined by haematoxylin-eosin (HE) staining. Nuclear factor erythroid 2-related factor 2 (Nrf2) and haem oxygenase-1 (HO-1) levels were detected by immunohistochemical analysis, while myeloperoxidase (MPO) levels were detected by immunofluorescence. DHE staining was used to determine reactive oxygen species (ROS) levels. Results The 72h survival rates of the polyHb and polyCOHb groups were 50.00% (6/12) and 58.33% (7/12) respectively, which were significantly higher than that of the 8.33% (1/12) in the HES group (p < 0.05). At RS 0h and RS 1h, the HbCO content of rats in the polyCOHb group (1.90 ± 0.21, 0.80 ± 0.21) g/L were higher than those in the polyHb group (0.40 ± 0.09, 0.50 ± 0.12)g/L (p < 0.05); At RS 1h, the MDA (41.47 ± 3.89 vs 34.17 ± 3.87 nmol/ml) in the plasma, Nrf2 and HO-1 content in the colon of rats in the polyCOHb group were lower than the polyHb group. And the SOD in the plasma (605.01 ± 24.46 vs 678.64 ± 36.37) U/mg and colon (115.72 ± 21.17 vs 156.70 ± 21.34) U/mg and the MPO content in the colon in the polyCOHb group were higher than the polyHb group (p < 0.05). Conclusions In these haemorrhagic shock/resuscitation models, both polyCOHb and polyHb show similar therapeutic effects, and polyCOHb has more effective effects in maintaining MAP, correcting acidosis, reducing inflammatory responses than that in polyHb.