Some studies accomplished thus far have indicated that mammographic patterns may serve as risk indices for breast cancer. The present investigation was undertaken to determine whether patterns are familial. Mammograms from 110 mothers and their daughters and 122 sister pairs and a like number of unrelated controls matched for patient age, reproductive history, and personal and family history of breast cancer and drawn from the files of the Radiology Department of Hutzel Hospital, Detroit, Michigan, were compared. A familial influence was indicated by the finding that pattern similarities were significantly stronger in test cases than in unrelated controls. This influence appeared to be manifested by premature age changes in daughters and younger sisters and consisted, primarily, of a decrease in DY and an increase in P2 patterns. The data suggest that performing mammography on first degree relatives of women with high risk P2 and DY patterns should be considered.