A positive family history is used in clinical practice as an indication of increased melanoma risk, yet there are no data on the accuracy of reported family histories of mel anoma. The validity of case-reported family history of melanoma was assessed in the course of a family and twin study of melanoma in Queensland, Australia, con ducted among the families of 2,118 melanoma cases diagnosed in Queensland between 1982 and 1990. A total of 913 melanoma cases made 1,267 reports of melanoma among their first-degree relatives. A total of 1,040 of these reports were checked, first through relatives them selves and then, if the relative also said they had had melanoma, through the relative's medical records. Med ical confirmation of melanoma as the diagnosis was ob tained for 623 reports (59.9%; 95% confidence interval 56.9-62.9): a false-positive reporting rate by cases of 40.1%. The level of false-positive reporting was lower for cases under 70 years of age, for women, for cases whose own diagnosis of melanoma was more than 5 years earlier, and for cases with three or more relatives with melanoma. Media campaigns in Queensland aimed at Increasing skin cancer awareness, and confusion be tween melanoma and other more common actinic neo plasms (basal and squamous cell carcinomas), may partly explain the high false-positive reporting rate ob served here. For this reason, it is difficult to generalize these findings to northern hemisphere populations where skin cancer is not such an Important public health issue.