Objective To establish a high-efficiency and stable mouse model of liver orthotopic transplantation for hepatocellular carcinoma. Methods A 5 μL solution containing 5 × 10⁵ H22 cells was injected into the left liver lobe using a 10 μL micro-syringe. In the key step of sealing the pinholes on the liver surface, two method were used to prevent cell leakage: cotton-pressed and heat-sealed. The size, weight, and pathology of liver tumors were examined on the 19th day after inoculation in preparation for establishing a stable and uniform liver orthotopic transplantation model. Results The liver tumor formation rate of both method was 100%. Although there was no statistically significant difference in tumor size, weight, or pathological performance between the two groups, the size and weight of the transplanted liver tumors were more uniform and showed smaller fluctuation in the heat-sealed than cotton-pressed group. In particular, significantly high rates of ascites (36. 4%) and abdominal tumors (36. 4%) were found in the cotton-pressed group (P< 0. 05). However, these abnormalities did not develop in the heat-sealed group. This finding indicates that the tumor cells more easily leaked from the pinholes in the cotton-pressed group, which may have been the main cause of the larger fluctuations in the size and weight of orthotopic transplanted liver tumors. In addition, one mouse with multiple abdominal tumors in the cotton-pressed group also exhibited tumor nodules, which may have been the result of metastasis from abdominal or ascitic tumor cells through the vasculature. Conclusions In the present study, we established a method for preparing a uniform and stable mouse model of liver orthotopic transplantation for hepatocellular carcinoma and described the key implementation points. The merits of short time, an easy operation, and stability provide a good reference for liver cancer research.