The traditional approach to the study of tumor markers has been based on the detection of substances either induced by or released by cancer cells. In high-risk patients with melanoma, we and others {1-4) have evaluated several potential tumor markers including lipid-associated sialic acid (in plasma), neuron-specific enolase, and circulating SI00 protein (a marker of neuroendocrin e tumors). Preliminary results using these markers have indicated that they have limited value in the clinical setting. The development of new molecular biology techniques, particularly the polymerase chain reaction (PCR), has provided a means by which molecular markers present at low copy numbers can be detected with a sensitivity significantly higher than that achieved using antibody-based techniques. In 1991, Smith et al. (5) proposed for the first time that melanoma cells could be detected in the peripheral blood using coupled reverse-transcription (RT) and PCR (RT-PCR) to target tyrosinase messenger RNA (mRNA). Tyrosinase, the first enzyme in melanin biosynthesis, is a mono-oxygenase that catalyzes the conversion of tyrosine to dopa and of dopa to dopaquinone. Thus, tyrosinase is one of the most specific markers of melanocytic differentiation. Tyrosinase expression can be detected not only in melanocytes and melanoma cells but also in Schwann cells (6). Nonetheless, because these cell types are not known to circulate, the detection of tyrosinase transcripts in the peripheral blood of patients with melanoma has been considered evidence of circulating melanoma cells. Since the original report from Smith et al. (5), several other groups have pursued this lead (7-11). The results are summarized in Table 1. As clearly shown in the table, the results are extremely variable. The most striking discrepancies were seen in patients with stage IV disease, in whom the sensitivity rates ranged from 0% to 100%. For instance, in the series of Brossart et al. (7), 100% of the patients with stage IV disease had marker-positive RTPCR findings regardless of response to therapy, whereas in the current report by Kunter et al. (12), only 28% of the patients with stage IV disease had positive RT-PCR. Furthermore, in the series of Kunter et al., positive RT-PCR was observed only in patients whose disease progressed during therapy. Significant discrepancies were also observed in patients with stage I and stage II disease, for whom the positive RT-PCR frequency ranged from 0% to 71%. Interestingly, the expected disease