Investigations of breast carcinogenesis often rely upon comparisons between cancer tissue and nonmalignant breast tissue. It is unclear how well common reference sources of nonmalignant breast tissues reflect normal breast tissue. Breast tissue samples were evaluated from three sources: (1) normal donor tissues in the Susan G. Komen for the Cure Tissue Bank at Indiana University Simon Cancer Center (KTB), (2) women who underwent reduction mammaplasty (RM) at Mayo Clinic Rochester, and (3) the Mayo Clinic Benign Breast Disease Cohort Study (BBD). Samples were examined histologically and assessed for proliferative disease and degree of lobular involution. Univariate comparisons were performed among the study groups, and multivariate analyses were performed with logistic regression to assess the association between study group and the presence of epithelial proliferative disease and complete lobular involution. Histologic data were collected for 455 KTB samples, 259 RM samples, and 319 BBD samples. Histologic findings and the frequency of epithelial proliferation were significantly different among the groups. Histologic abnormalities were seen in a minority of the KTB samples (35%), whereas an abnormality was present in 88% of RM tissues and 97.5% of BBD samples. The presence of proliferative disease (with or without atypical hyperplasia) was present in 3.3% of normal donors (3.3%), 17% of RM samples, and 34.9% of BBD samples (P < 0.0001 for each comparison). Multivariate analyses confirmed that these differences remained significant and also showed higher likelihood of complete lobular involution in the normal donor samples compared to RM and BBD tissues. Compared to benign breast disease tissues and reduction mammaplasty tissues, breast tissue samples from normal donors have significantly fewer histologic abnormalities and a higher frequency of more complete lobular involution. Breast tissue samples from normal donors represent a unique tissue resource with histologic features consistent with lower breast cancer risk.