We conducted genetic linkage analyses of breast cancer in 20 pedigrees, each having at least one case of bilateral breast cancer diagnosed before 50 years of age. We tested for linkage using inheritance models from previous segregation analyses, incorporating differences in risk based on menopausal status into the analyses. We tested for heterogeneity by predividing the data set based on the interval between diagnoses of the proband's two primaries (less than 1 year (synchronous) versus at least 2 years (asynchronous], and on the histological types of breast cancer in the pedigrees. Very tight linkage could be excluded between breast cancer and ABO, GC, GPT, MNS, and PGM1 for some of the different linkage analyses. A maximum lod score of +1.01 (at theta = 0.001) between ACP1 and a breast cancer susceptibility locus was seen in the asynchronous all-cases subsample.