Genetic predisposition plays a key role in the development of familial breast cancer. In spite of strong familial clustering of the disease and extensive efforts made during the past decade; however, progress has been slow in identifying genetic predisposition for the majority of familial breast cancer families. The question arises therefore as to whether current approaches are adequate in identifying the unknown genetic predisposition. We analyzed eight members of a BRCA1‐, BRCA2‐, p53‐, and PTEN‐negative breast cancer family, of which five had breast cancer, one is an obligate gene carrier, and two were unaffected. We sequenced the entire coding region of the genome for each member using exome sequencing to identify nonsynonymous variants. We identified 55 nonsynonymous germline variants affecting 49 genes in multiple members of the family, of which 22 are predicted to have damaging effects. We validated 20 of the 22 selected variants in the family by Sanger sequencing. Two variants in KAT6B, an acetal transferase gene, were identified in six family members of which five were affected with breast cancer and one is the unaffected obligate carrier. We further examined the presence of the identified variants in a cohort of 40 additional breast cancer cases from 22 familial breast cancer families, but none of the 22 variants was detected in these cases. Sequencing the entire coding exons in KAT6B detects no variants in these cases. Our results show that genetic predisposition for familial breast cancer can be rich in an affected family, but the predisposition can be family‐specific. As such, it will be difficult to detect them by applying population‐based approach. Our study supports the concept that focusing on each affected family will be required to determine the genetic predisposition for many familial breast cancer families whose genetic dispositions remain unknown.