Objective The established link between oestrogen and breast cancer occurs via both oestrogen receptor (ER)-mediated and non ER-mediated mechanisms. The term genotoxic estrogens describes mutagenic metabolites, including oestrogen catechols and quinones, which have been linked to breast carcinogenesis in post-menopausal women. We aimed to assess whether the route of administration of 17 beta oestradiol (E-2) affects the accumulation of genotoxic oestrogen metabolites in a model of ovarian failure in young girls with Turner syndrome. Methods Stored plasma samples obtained at 0 and 12 months were used from 40 adolescents with Turner syndrome who participated in a 12 months randomized controlled trial of the metabolic impact of E-2 orally (2 mg/d) vs transdermally (100 mu g/d); dose escalation allowed matching of unconjugated E-2 levels in the parent study. We measured 12 oestrogen metabolites (total concentrations = conjugated and unconjugated) using a highly sensitive LCMSMS assay. Results from 48 normally menstruating adolescents were used for comparison. Results After treatment, least square mean (SE) total E-2 concentrations were higher in the oral vs transdermal group (6784 pmol/L vs 1123 [1614], P < 0.0001), as was oestrone (E-1) (91 060 pmol/L vs 19 278 [16 534], P < 0.0001). Also, higher after oral treatment were catechol-oestrogens 4-hydroxy-E-2 (149 vs 28 [+/- 49] pmol/L), 2-hydroxy-E-2 (300 vs 76 [+/- 52]), 4-hydroxy-E-1 (450 vs 105 [+/- 113]), 2-hydroxy-E-1 (3094 vs 740 [+/- 684]) and 16 alpha-hydroxy-E-1 (3,007 vs 157 [+/- 534]) (<0.001 between groups). Levels were much closer to controls in the transdermal group. Conclusions Common feminizing doses of oral oestradiol for 12 months result in substantial accumulation of unphysiologic, genotoxic oestrogens compared to transdermal oestradiol, expanding concerns about oral oestrogens' first hepatic passage. Further studies assessing long-term risks of these metabolites in women taking different forms of oestrogen are needed.