Previous studies evaluating pregnancy hormone levels and maternal breast cancer were limited to surrogate indicators of exposure. This study directly evaluates the association between measured serum steroid hormone levels during pregnancy and maternal risk of breast cancer. A nested case-control study was conducted to examine third-trimester serum levels of total unconjugated estradiol, estrone, estriol, and progesterone in women who were pregnant between 1959 and 1966. Cases (n = 194) were diagnosed with in situ or invasive breast cancer between 1969 and 1991. Controls (n = 374) were matched to cases by age at the time of index pregnancy, using randomized recruitment. Elevated progesterone levels were associated with a decreased incidence of breast cancer [odds ratio (OR) for progesterone ≥270 ng/ml, 0.49; 95% confidence interval (CI), 0.22–1.1] relative to those below the lowest decile. This association was stronger for cancers diagnosed at or before age 50 (OR for progesterone ≥270 ng/ml, 0.3; 95% CI, 0.1–0.9). Increased estrone levels were associated with an increased incidence overall (OR for estrone ≥18.7 ng/ml, 2.5; 95% CI, 1.0–6.2), whereas a positive association with estradiol was not observed. Too few cases occurred within 15 years of the index pregnancy to compare adequately the short- and long-term effects of pregnancy hormone exposure. When estrogen-to-progesterone ratios were evaluated, there was an indication of a modest increased incidence of breast cancer for those with high total estrogens and high estrone levels relative to progesterone. These findings suggest that pregnancy steroid hormone levels are risk factors for breast cancer.