Abstract：Objective：To analyze the medication rule of Chinese medicine for pancreatic cancer based on data mining. Methods：The prescriptions of Chinese medicine for pancreatic cancer were retrieved from the Dictionary of Tumor Prescriptions and the Complete Compendium of Tumor Prescriptions； the frequency analysis，association rule analysis，cluster analysis and principal component factor analysis were used to summarize the medication rules of Chinese medicine for pancreatic cancer. Results： A total of 161 prescriptions were included， involving 263 Chinese medicinals. The high-frequency medicines of Chinese medicine for pancreatic cancer include Poria，Hedyotis Diffusae Herba，Angelicae Sinensis Radix， Artemisiae Scopariae Herba，Curcumae Radix，etc.. The therapeutic medicines are mainly those for clearing heat，invigorating blood and dissolving stasis，and supplementing deficiency；the main medicinal nature is cold and warm； the main medicinal flavor is bitter and acrid， mainly attributed to liver and stomach meridians. The core medicine combinations like Curcumae Radix-Salviae Miltiorrhizae Radix et Rhizoma， Poria-Atractylodis Macrocephalae Rhizoma，and Curcumae Radix-Salviae Miltiorrhizae Radix et Rhizoma+ Poria were obtained by association rule analysis；seven new prescriptions，such as those for breaking up blood stasis and resolving masses，activating blood，resolving stasis and relieving pain，soothing the liver， regulating qi and fortifying the spleen， supplementing the center and boosting qi， and clearing heat and detoxifying were obtained by cluster analysis；eight common factors were obtained by principal component factor analysis. Conclusion：The treatment of pancreatic cancer with Chinese medicine is mainly based on clearing heat and resoling toxins， activating blood and resolving stasis， supplemented by medicines for supplementing deficiency， draining dampness， and boosting qi and fortifying the spleen， etc.. In clinical practice， Yinchenhao Decoction， Shenling Baizhu Powder and Gexia Zhuyu Decoction are often selected for syndrome differentiation and treatment，with modification based on different symptoms.