Objective:To evaluate the reproducibility of dynamic CT perfusion(CTP)imaging for liver cancer. Methods:Multi-slice spiral CT(Siemens sensation 64,Germany)was used and the data was processed with liver perfusion software-Body PCT,for which the mathematical model of slope and the analysis method of Patlak was used. First,a relative coincident and standard post-processing method was established. CT perfusion image was obtained after software processing,and each ROI was marked out with mouse,including the size of the tumor and the half-diameter ROI of the tumor. Then a post-processing method for comparison was built up:①To observe the reproducibility of the same operator:The operators who have different skills and experience(A and B)separately made the post-processing of CTP. The post-processing results from the same operator were statistically analyzed and the consistency was evaluated. ②To observe the reproducibility between the two operators:The results from the two operators were compared to evaluate the effects of different operators on the reproducibility of CTP. Results:The result from operator A was consistent with that from B for the first time; and there was also a good consistency between the results from the same operator. Conclusion:Dynamic CT perfusion imaging in liver cancer is a reproducible method.