Polycystic ovary syndrome (PCOS) affects 5% of reproductive aged women and is the leading cause of anovulatory infertility. A hallmark of PCOS is excessive theca cell androgen secretion, which is directly linked to the symptoms of PCOS. Our previous studies demonstrated that theca cells from PCOS ovaries maintained in long term culture persistently secrete significantly greater amounts of androgens than normal theca cells, suggesting an intrinsic abnormality. Furthermore, previous studies suggested that ovarian hyperandrogenemia is inherited as an autosomal dominant trait. However, the genes responsible for ovarian hyperandrogenemia of PCOS have not been identified. In this present study, we carried out microarray analysis to define the gene networks involved in excess androgen synthesis by the PCOS theca cells in order to identify candidate PCOS genes. Our analysis revealed that PCOS theca cells have a gene expression profile that is distinct from normal theca cells. Included in the cohort of genes with increased mRNA abundance in PCOS theca cells were aldehyde dehydrogenase 6 and retinol dehydrogenase 2, which play a role in all-trans-retinoic acid biosynthesis and the transcription factor GATA6. We demonstrated that retinoic acid and GATA6 increased the expression of 17alpha-hydroxylase, providing a functional link between altered gene expression and intrinsic abnormalities in PCOS theca cells. Thus, our analyses have 1) defined a stable molecular phenotype of PCOS theca cells, 2) suggested new mechanisms for excess androgen synthesis by PCOS theca cells, and 3) identified new candidate genes that may be involved in the genetic etiology of PCOS.