The development and progression of cancer is accompanied by complex changes in patterns of gene expression. In the USA, malignant neoplasms of the skin are the most common cancers of humankind. Incidence rates of melanoma have risen especially steeply since the mid-1970s with an almost complete absence of significant advances in non-surgical treatment of advanced malignant melanoma over this time period. Efforts to reduce mortality must rely on earlier diagnosis, but also on understanding the biology and genetics of this difficult disease. We have used high-density microarrays to search for differences in gene expression profiles associated with melanoma. Applications of multiple data analysis and integration methods including multidimensional scaling to represent the relationships among cell lines and tumor biopsies will be presented in order to document a consistent pattern of gene expression characterizing a subset of these cancers. This information will be related to clinical and known biologic information. These data are providing the leads for further investigation of the genetic basis of the tumorigenic phenotype of melanocytic lesions.