Eleven children 10 years of age or younger (6 of them below the age of 5) with primary cutaneous melanoma constituted our series. All of the melanomas occurred de novo and all metastasized; one child has already died. In no instance was melanoma a consideration clinically and in none was a diagnosis of melanoma made straightforwardly by the histopathologist who first "signed it out." Despite the inability of clinicians and pathologists to diagnose correctly, with repeatability, melanomas that develop in children yet to be pubescent, those neoplasms, nonetheless, are melanomas and, therefore, criteria employed currently for diagnosis of melanoma, clinically and histopathologically, must be improved in order that they be applicable equally to melanomas in pre- and post- pubescents. The vaunted ABCDs (Asymmetry, Border irregularity, Color variability, Diameter >6.0 mm) surely do not work for melanomas that appear in children who are prepubescent. Moreover, melanomas that occur in these children have distinctly different architectural and cytopathological features from those that arise in post-pubescents, often being confused as they are by conventional microscopy with a Spitz"s nevus. As a rule, melanomas in prepubescent children grow much more rapidly then those in adults but, like them, have the capability to disseminate widely and cause death.