The diagnosis of polycystic ovary syndrome (PCOS) remains challenging due to limited data regarding normative cut-offs for the diagnostic features in different subpopulations. We aim to conduct a systematic review, build a comprehensive repository of de-identified individual participant data (IPD), and define normative ranges and diagnostic cut-offs for all PCOS diagnostic features. We will conduct a systematic search of MEDLINE and EMBASE databases for studies that assessed PCOS diagnostic features in unselected women. Two reviewers will assess eligibility and perform quality appraisal. Authors of included studies will be invited to contribute IPD. Primary variables include directly assessed modified Ferriman Gallwey (mFG) scores; menstrual cycle lengths; follicle number per ovary (FNPO), ovarian volume (OV), anti-Müllerian hormone (AMH); circulating androgens, including total testosterone (TT), free testosterone, bioavailable testosterone, free androgen index (FAI), androstenedione (A4), and dehydroepiandrosterone sulphate (DHEAS). Normative ranges and cut-offs will be defined using cluster analysis. Monash University Human Research Ethics Committee granted ethical approval (26938/0 1/12/2020), all IPD will be de-identified and primary studies have ethical approval from their institutional ethics committees. Findings will clarify distinction between PCOS and non-PCOS populations, and inform the update of the international evidence-based guidelines for the assessment and management of PCOS.