Epidemiological studies have demonstrated that men with a family history of prostate cancer are at an increased risk for this disease. This important observation has led a number of research teams, including our own, to collect DNA samples and clinical data from prostate cancer families, with the goal of localizing and characterizing prostate cancer susceptibility genes. The candidate tumor suppressor gene PTEN (also called MMAC1) has recently been shown to be somatically altered in several common malignancies, including cancers of the brain, kidney, skin, thyroid, endometrium, breast, and prostate. Germ-line mutations in this gene, which maps to chromosome 10q23, have been associated with Cowden disease, an autosomal dominant cancer predisposition syndrome that is characterized by multiple hamartomas. Although prostate cancer is not typically associated with Cowden disease, previous studies of sporadic prostate cancers demonstrate loss of heterozygosity at 10q23 loci in ∼25% of cases. We, therefore, hypothesized that germ-line mutations in the PTEN gene may predispose to prostate cancer in a subset of families, particularly those in which cancers of the breast, kidney, and/or thyroid also segregate. To test this hypothesis, DNA was isolated from whole blood of 11 prostate cancer patients from 10 unrelated families. Four of the 10 families met the previously established clinical criteria for hereditary prostate cancer. Eight of the 11 men had at least one second primary malignancy, including cases of neuroendocrine cancer, glioblastoma multiforme, melanoma, kidney, and thyroid cancer. Although we identified some common as well as some unique polymorphisms, no nonsense or missense mutations were identified in any of the 11 samples. To further examine the possibility that PTEN mutations contribute to prostate cancer predisposition, we also studied the probands from each of 10 families with early-onset and/or multiple individuals with prostate cancer. Sequence analysis of the PTEN gene in these 10 men also revealed no mutations or novel polymorphisms. We conclude that germ-line mutations in the PTEN are unlikely to contribute in a significant way to the inherited predisposition to prostate cancer.