High-dose estrogen exposure during adolescence has been hypothesized to increase a woman's breast cancer risk, possibly mediated through an increase in mammographic density, a well-established breast cancer risk factor. In 2006 to 2007, we conducted a retrospective study of women assessed for tall stature as an adolescent between 1959 and 1993. Eligible participants were ages > or =40 years and treated during adolescence with 3 mg diethylstilbestrol or 150 microg ethinyl estradiol daily or untreated. Mammograms from 167 treated and 142 untreated women were digitized. Total breast area, dense area, nondense area, and percent density were measured using a computer thresholding technique. Data on potential determinants were collected from medical records and telephone interview. Treated women had, on average, 17% lower dense area (P = 0.032). Means (95% confidence intervals) adjusted for age and body mass index for treated and untreated women were 24.5 cm(2) (21.8-27.2) and 29.1 cm(2) (26.0-32.4), respectively. There was no difference in adjusted means (95% confidence intervals) between treated and untreated women for nondense area [71.7 cm(2) (66.2-77.7) versus 70.5 cm(2) (64.7-76.9); P = 0.78], percent dense area [24.8% (22.4-27.4) versus 27.7% (24.8-30.7); P = 0.16], or total area [105.6 cm(2) (100.1-111.4) versus 109.3 cm(2) (103.1-115.8); P = 0.41], respectively. High-dose estrogen exposure during adolescence appears to curtail growth of mammographically dense tissue and therefore is unlikely to increase breast cancer risk through mechanisms related to mammographic density.