ABSTRACTResearch questionPolycystic ovary syndrome (PCOS) is a common endocrine disorder with unclear etiology. Are there any genes that are pleiotropically or potentially causally associated with PCOS?DesignWe applied the summary data-based Mendelian randomization (SMR) method integrating genome-wide association study (GWAS) for PCOS and expression quantitative trait loci (eQTL) data to identify genes that were pleiotropically associated with PCOS. We performed separate SMR analysis using eQTL data in the ovary and whole blood.ResultsAlthough no genes showed significant pleiotropic association with PCOS after correction for multiple testing, some of the genes exhibited suggestive significance. RPS26 showed the strongest suggestive pleiotropic association with PCOS in both SMR analyses (β[SE]=0.10[0.03], PSMR=1.72×10−4 for ovary; β[SE]=0.11[0.03], PSMR=1.40×10−4 for whole blood). PM20D1 showed the second strongest suggestive pleiotropic association with PCOS in the SMR analysis using eQTL data for the whole blood, and was also among the top ten hit genes in the SMR analysis using eQTL data for the ovary. Two other genes, including CTC-457L16.2 and NEIL2, were among the top ten hit genes in both SMR analyses.ConclusionWe identified multiple genes that were potentially involved in the pathogenesis of PCOS. Our findings provided helpful leads to a better understanding of the mechanisms underlying PCOS, and revealed potential therapeutic targets for the effective treatment of PCOS.KEY MESSAGEPolycystic ovary syndrome (PCOS) is a common endocrine disorder, and its etiology can be multifaceted. We found that multiple genes were potentially involved in the pathogenesis of PCOS. The findings revealed the genetic mechanisms underlying PCOS and potential therapeutic targets for the effective treatment of PCOS.