Background: Many animal experimental studies showed that abdominal tumor cells will be widely spread during laparoscopic treatment and grow into metastases. These results are different from clinical observations. There is a hypothesis that too much tumor cells was injected in the animals lead to the results of theses bias. We aim to learn the difference of abdominal cavity volume between human body and the nude mice and to determine reasonable amount of tumor cells in the animal experiments. Methods: The insufflated CO2 volume which represents the capacity of the abdominal cavity was recorded during laparoscopic process in 212 patients and 20 nude mice respectively, the relative volume of nude mice and human body was calculated.Based on data from the literature and this study , the amount of tumor cells in the animal experiments was determined.According to these data, we set up a new animal model and a traditional one respectively,and compared the rate of successful modeling and tumor formation between two animal models. Results: The intraperitoneal volumes of humans and nude mice were 3.01±0.36 L and 0.011±0.001 L respectively.The number of tumor cells that be uesd in animal should be approximately 0.26×105 in terms of known data in human beings.Compared with the traditional animal model which formed a large number of intraperitoneal tumor metastasis, the new animal model was shows more moderately,and the rate of successful modeling was similar. Conclusion: In animal experiments, to simulate the clinical situation, about 0.26×105 tumor cells should be inject in peritoneal cavity of the nude mice. Creative Commons Attribution License