By use of the decision tree algorithm and diagnostic indexes, the paper sets up the discrimination rules to make differential diagnosis of Primary Hepatocelluar Carcinoma(PHC) based on basic data of 95 patients with PHC and 190 patients with liver cirrhosis, including the CT diagnosis, testing results of imaging and serologic markers such as the HbsAg, AFP, CEA and AFU, sex and age, etc. As indicated by the results, the data mining technology represented by the decision tree can support the differential diagnosis of PHC.