Objective: To evaluate the feasibility of establishing portal vein thrombosis (PVT) in rats by repeated interruption of the portal vein. Methods: Ninety SD rats were randomly and equally divided into experimental group and control group. Rats in experimental group underwent repeated ligation of the portal vein for induction of PVT, and those in control group underwent exposure of the portal vein after laparotomy. At preoperative and postoperative 30 min as well as postoperative 3, 6, 12 and 24 h, the PVT formation in the two groups of rats was observed by vascular Doppler ultrasound, and the levels of P-selectin, D-dimer and fibrinogen in the venous blood of the rats were determined. All rats were sacrificed 24 h after operation, and PVT formation was observed by gross examination and pathological examination was also performed. Results: In experimental group, 5 rats died, 6 rats did not have PVT formation and 34 rats (75.56%) developed stable PVT at 24 h after operation; in control group, 2 rats died, a small amount of mural thrombus was seen in 3 rats and no thrombosis was observed in the other 40 rats, and the difference in PVT formation had statistical significance between the two groups (P<0.05). The length of the thrombus in rats of experimental group was gradually increased with time prolongation. The variables associated with thrombosis showed no significant difference between the two groups at preoperative 30 min and postoperative 30 min and 3 h (all P>0.05), but they were significantly higher in experimental group than those in control group at each time point from postoperative 6 h (all P<0.05). Conclusion: Repeated interruption of the portal vein can generate a stable and clinical relevant PVT model in rats. P-selectin level, D-dimer content and fibrinogen content can be used as reference indexes.