To investigate the principle and regularity of acupoint selection in the treatment of post-stroke cognitive impairment (PSCI) with acupuncture.Methods:CKNI，WanFang Data，and VIP were searched for clinical studies of acupuncture in the treatment of PSCI published from January 1，2005，to December 1，2020.SPSS 22.0 and SPSS Modeler 18.0 were used for descriptive analysis，association analysis，and cluster analysis.Results:A total of 120 acupuncture prescriptions were included，involving 81 acupoints，with a frequency of 967.1) In terms of frequency，the top acupoints were Baihui (GV 20)，Shenting (GV 24)，Sishencong (EX-HN 1)，Sanyinjiao (SP 6)，and Zusanli (ST 36).2) Association rule analysis showed that Baihui (GV 20)-Shenting (GV 24) was the highest in the support degree (50.83%) with a confidence degree of 50.83%.3) The acupoints were classified into six categories according to the frequency≥10.4) In terms of meridian tropism of the selected acupoints，the top meridian was the governor vessel，followed by the three yang meridians of the foot.5) In terms of specific points，confluence points were most selected，and five transport points，source points，connecting points，and back transport points were also widely applied.6) In the treatment，acupuncture was commonly combined with rehabilitation training，western medicine，electro-acupuncture，and Chinese medicine.Conclusion:In the past 15 years，acupuncture treatment of PSCI was mainly applied to the acupoints on the governor vessel，such as Baihui (GV 20)，based on syndrome differentiation，and the acupoints regulating zang-fu organs and qi movement after compatibility were highlighted.In clinical treatment，the correlation regularity obtained by data mining and the results of cluster analysis can be used to select acupoints reasonably to improve the curative efficacy.