To investigate the medication regularity and core prescriptions of marketed Chinese patent medicines against diabetes by data mining in order to provide references for clinical treatment of diabetes and new ideas for new drug development.Methods:The Chinese Pharmacopoeia (2020) and Yaozh.com were searched for prescriptions of Chinese patent medicine against diabetes with key terms of “consumptive thirst” and “diabetes” and the database was established.Crude drugs involved were summarized and analyzed through Excel，TCM Knowledge Service and Sharing System，and Liquorice.Results:Seventy-one Chinese patent medicines were included，involving 158 crude drugs.The frequency of use，properties (nature，flavor，and meridian tropism)，and compatibility of these drugs were obtained.The core prescriptions of 71 crude drugs were screened out by complex network analysis.As revealed by syndrome classification of included prescriptions，the main syndromes included qi-yin deficiency，qi-yin deficiency and internal heat，yin deficiency and internal heat，qi-yin deficiency and blood stasis，and yin deficiency-generated internal heat combined with qi deficiency and blood stasis，and the corresponding core prescriptions were obtained.Conclusion:Through the analysis of prescriptions of Chinese patent medicines marketed for the treatment of diabetes by mathematical statistics methods，the medication regularity and core prescriptions can be obtained，which can provide ideas for clinical medication and the development of new drugs．