Background: Increasing incidence and mortality rates from cutaneous melanoma are a major public health concern. As part of a national effort to enhance early detection of melanoma/skin cancer, the American Academy of Dermatology (AAD) has sponsored an annual education and early detection program that couples provision of skin cancer information to the general public with almost 750,000 free skin cancer examinations (1985-1994). Objective: To begin to evaluate the impact of this effort, we determined the final pathology diagnosis of persons attending the 1992-1994 programs who had a suspected melanoma at the time of examination. Methods: We directly contacted all such persons by telephone or mail and received pathology reports from those who had a subsequent biopsy. Results: We contacted 96% of the 4458 persons with such lesions among the 282,555 screenings in the 1992- 1994 programs. We obtained a final diagnosis for 72%, and the positive predictive value for melanoma was 17%. Three hundred seventy-one melanomas were found in 364 persons. More than 98% had localized disease. More than 90% of the confirmed melanomas with known histology were in situ or 'thin' lesions (≤1.50 mm thick). The median thickness of all melanomas was 0.30 mm. The 8.3% of AAD cases with advanced melanoma (metastatic disease, regional disease, or lesions ≥1.51 mm) is a lower proportion than that reported by the 1990 Surveillance, Epidemiology and End Result Registry. The rate of thickest lesions (≥4 mm) and late-stage melanomas among all participants was 2.83 per 100,000 population. Of persons with a confirmed melanoma, 39% indicated (before their examination) that without the free program, they would not have considered having a physician examine their skin. Conclusion: The 1992-1994 free AAD programs disseminated broad skin cancer educational messages, enabled thousands to obtain a free expert skin cancer examination, and found mostly thin, localized stage 1 melanomas (usually associated with a high projected 5-year survival rate). Because biases impose possible limitations, future studies with long-term follow-up and formal control groups should determine the impact of early detection programs on melanoma mortality.