We have demonstrated statistical evidence for a major gene predisposing to cancer in kindreds of childhood and adolescent soft tissue and bone sarcoma patients. We have evidence for heterogeneity in risk by kindred, by generation, by sex of affected parent, and by selected proband characteristics. The familial pattern of cancer changes over time, with increasing evidence for a major gene evolving after more than 20 years of observation. To date, some of the kindreds that provide strong evidence for a major gene have been found to have germline mutations in the tumor suppressor gene, p53. We now propose to characterize the heterogeneity in risk, using a strategy of combined linkage and segregation analysis, with the additional critical years of observation that will accrue. Initially we will test for linkage to p53. However if the findings support involvement of other genetic loci, the strategy can be generalized to identify other major genes or modifying genes. The immediate goal will be to identify the extent to which germline p53 mutations can account for the observed familial cancer aggregation, to characterize the phenotype associated with p53 germline mutations, and to identify any variation in risk not attributable to p53 as a major gene. findings from this project should provide sufficient information to develop guidelines for genetic testing of childhood sarcoma patients, guidelines for genetic counseling regarding the implications of p53 germline mutations, and a framework from which to investigate the role of other genetic loci in familial cancer aggregates.