BackgroundPost-stroke hemiparesis strongly affects stroke patients’ activities of daily living and health-related quality of life. Scalp acupuncture (SA) is reportedly beneficial for post-stroke hemiparesis. However, there is still no standard of SA for the treatment of post-stroke hemiparesis. Apriori algorithm-based association rule analysis is a kind of “if-then” rule-based machine learning method suitable for investigating the underlying rules of acupuncture point/location selections. This study aimed to investigate the core SA combinations for the treatment of post-stroke hemiparesis by using a systematic review and Apriori algorithm-based association rule analysis.MethodsWe conducted a systematic review to include relevant randomized controlled trial (RCT) studies investigating the effects of SA treatment in treating patients with post-stroke hemiparesis, assessed by the Fugl-Meyer Assessment (FMA) score. We excluded studies using herbal medicine or manual acupuncture.ResultsWe extracted 33 SA locations from the 35 included RCT studies. The following SA styles were noted: International Standard Scalp Acupuncture (ISSA), WHO Standard Acupuncture Point Locations (SAPL), Zhu’s style SA, Jiao’s style SA, and Lin’s style SA. Sixty-one association rules were investigated based on the integrated SA location data.ConclusionsSAPL_GV20 (Baihui), SAPL_GV24 (Shenting), ISSA_MS6_i (ISSA Anterior Oblique Line of Vertex-Temporal, lesion-ipsilateral), ISSA_MS7_i (ISSA Posterior Oblique Line of Vertex-Temporal, lesion-ipsilateral), ISSA_PR (ISSA Parietal region, comprised of ISSA_MS5, ISSA_MS6, ISSA_MS7, ISSA_MS8, and ISSA_MS9), and SAPL_Ex.HN3 (Yintang) can be considered the core SA location combination for the treatment of post-stroke hemiparesis. We recommend a core SA combination for further animal studies, clinical trials, and treatment strategies.