Abstract: OBJECTIVE To explore the regularity of Chinese patent medicine to treat dementia by using data mining technology, and to provide reference for guiding the development of clinical medicine and Chinese patent medicine. METHODS Selected and sorted out the treatment of dementia included in the “National Standards of Traditional Chinese Medicines”, “Ministry of Health of the People's Republic of China Drug Standards·Traditional Chinese Medicine Preparations”, 2010 edition “Instructions for Clinical Use of Traditional Chinese Medicine Preparations”, 2015 edition “Chinese Pharmacopoeia”. For finished drugs, the drugs that meet the inclusion criteria were entered into an Excel spreadsheet database for frequency analysis, and the data were analyzed by SPSS Statistics 23.0 and SPSS Modeler 14.1 statistical software. RESULTS A total of 183 Chinese patent medicines were included, and the contained medicines had 421 flavors. The Chinese medicines with higher frequency were radix rehmanniae, angelica, poria, wolfberry, ginseng. Efficacy classification mainly focused on tonic drugs, four natures and five flavors were mainly sweet and warm, and the meridian mainly focused on the five internal organs. The association rule analysis revealed that there were 7 drug pair combinations with the highest association strength. Six common factors were extracted from the factor analysis. Four clusters were obtained from the systematic cluster analysis. CONCLUSION The regularity of Chinese patent medicines for the treatment of dementia shows the common features of nourishing the kidney and promoting essence, promoting blood circulation and removing phlegm.