Natural products are widely promoted to increase or suppress androgen activity, but the strength and clinical relevance of the evidence vary markedly by preparation, population, endpoint, and study design. We conducted a structured critical review of human and mechanistic studies of androgen-modulating botanicals, with separate appraisal of in silico predictions, biochemical or cellular target effects, hormonal biomarkers, clinical outcomes, extract characteristics, and safety. Standardized extracts of Withania somnifera, Trigonella foenum-graecum, and Eurycoma longifolia have limited-to-moderate evidence for modest increases in circulating testosterone in selected male populations, particularly men with low baseline testosterone, aging-related symptoms, infertility, or stress; evidence for durable symptomatic or functional benefit is less consistent and cannot be generalized across extracts. Mentha spicata tea has small, short-duration human studies showing reductions in free or total testosterone in women with polycystic ovary syndrome or hirsutism, but objective improvement in hirsutism has not been demonstrated. Curcumin has preclinical evidence for effects on steroidogenic enzymes and androgen-receptor pathways, whereas human studies predominantly report metabolic outcomes in polycystic ovary syndrome or prostate-specific antigen-related endpoints rather than established systemic anti-androgenic efficacy. Dihydrotanshinone I from Salvia miltiorrhiza has biochemical evidence for preferential inhibition of CYP17A1 17,20-lyase and tanshinones suppress AR signaling in preclinical models, but no controlled human study has established androgen-lowering efficacy. Serenoa repens illustrates the frequent disconnect between plausible 5α-reductase inhibition and largely negative trials for benign prostatic hyperplasia. Molecular docking is treated as hypothesis-generating rather than confirmatory. Evidence grades are preparation- and population-specific and do not apply to uncharacterized products. Clinical translation requires chemically characterized preparations, population-specific trials, clinically meaningful endpoints, risk assessment, and active surveillance for hepatotoxicity, cardiovascular effects, contamination, and drug interactions.
