Although it is well known that fibroblasts can overgrow melanoma cells in culture during early passaging, some investigators have claimed recently that these fibroblast-like cells are actually melanocytes de-differentiated in vitro. In order to test these claims, we examined 27 melanoma-derived cell lines using anti-melanoma and anti-fibroblast monoclonal antibodies, the dioxyphenylalanine cytochemical reaction specific for melanocytes, the leucine aminopeptidase cytochemical reaction for fibroblast cells, and chromosomal analysis. All these assays - which discriminate melanomas from fibroblasts - confirmed our initial classification by morphology, namely that thirteen of the lines were melanocytic and twelve were fibroblastic, while the remaining two were a mixture of the two cell types. There was no indication of a mixed phenotype which would have been the result of some transaction in the state of differentiation. Our results suggest that the 27 cell lines and cultures are composed of pure melanocytic populations, or are fibroblasts as a result of fibroblast overgrowth. The two cultures containing a mixed cell population are those in which one cell type has not overtaken the other. There is no evidence to support the idea that some true melanoma cultures have the morphology of fibroblast-like cells.