Melanoma may cluster in families with 'family cancer syndromes' in which there is a predisposition to a variety of different tumours. Other families seem vulnerable to melanoma alone. In the majority of these families, the propensity to melanoma is associated with the presence of abnormal melanocytic naevi, the so-called atypical mole syndrome (AMS) phenotype. However, in a smaller number of families, individuals are susceptible to melanoma but have normal naevi. There appears, therefore, to be clinical (and probably genetic) heterogeneity. Segregation analysis does not support a predisposition by single dominant gene as an explanation for the AMS/melanoma syndrome. To date, a single gene which is clearly important for susceptibility to melanoma has not been identified. Karyotypic studies of melanoma tumours have pointed to chromosomes 1, 6, 7, 9 and 10 as possible sites for melanoma related genes. Loss of heterozygosity studies have suggested that chromosome 9 may carry a tumour suppressor gene important in familial disease, and linkage studies appear to confirm this. It is not yet clear, however, what percentage of familial melanoma is attributable to this gene. A more longstanding suggestion that a gene on chromosome 1 may be important has not been confirmed, but a chromosome/gene may be responsible for susceptibility in a small subset of melanoma families. Even within AMS families, there is a lack of concordance between the AMS phenotype and susceptibility to melanoma. This might be explained either by the effects of modifying genes, or the environment.