IMPORTANCE:After menopause, estradiol (E2) is predominately an intracrine hormone circulating in very low serum concentrations. OBJECTIVE:The objective of this work is to examine determinants of E2 concentrations in women beyond age 70 years. DESIGN AND SETTING:A cross-sectional, community-based study was conducted. PARTICIPANTS:A total of 5325 women participated, with a mean age of 75.1 years (± 4.2 years) and not using any sex steroid, antiandrogen/estrogen, glucocorticoid, or antiglycemic therapy. MAIN OUTCOME MEASURES:Sex steroids were measured by liquid chromatography-tandem mass spectrometry. Values below the limit of detection (LOD; E2 11 pmol/L [3 pg/mL] were assigned a value of LOD/√2 to estimate total E2. RESULTS:E2 and estrone (E1) were below the LOD in 66.1% and 0.9% of women, respectively. The median (interdecile ranges) for E1 and detectable E2 were 181.2 pmol/L (range, 88.7-347.6 pmol/L) and 22.0 pmol/L (range, 11.0-58.7 pmol/L). Women with undetectable E2 vs detectable E2 were older (median age 74.1 years vs 73.8, P = .02), leaner (median body mass index [BMI] 26.8 kg/m2 vs 28.5, P < .001), and had lower E1, testosterone and DHEA concentrations (P < .001). A linear regression model, including age, BMI, E1, and testosterone, explained 20.9% of the variation in total E2, but explained only an additional 1.2% of variation over E1 alone. E1 and testosterone made significant contributions (r2 = 0.162, P < .001) in a model for the subset of women with detectable E2. CONCLUSIONS:Our findings support E1 as a principal circulating estrogen and demonstrate a robust association between E1 and E2 concentrations in postmenopausal women. Taken together with prior evidence for associations between E1 and health outcomes, E1 should be included in studies examining associations between estrogen levels and health outcomes in postmenopausal women.