Introduction: Lymph node status in patients with cutaneous malignant melanoma is the most important prognostic factor. Patients with clinically positive nodes (stage III) should undergo therapeutic lymphadenectomy; however, the surgical approach to the regional disease in patients with negative clinical examination (stage I and II) is still controversial. Selective lymphadenectomy consists of the intraoperative identification of the first node in the nodal basin, the sentinel lymph node (SLN). Routine examination, serial sectioning, and immunohistochemistry may underestimate the presence of tumor cells. PCR is a molecular biology technique that may be useful for the detection of malignant melanoma nodal metastases in the SLN. Aim: The aim of this study was to use tyrosinase messenger RNA (mRNA) amplification for the detection of micrometastases in fresh frozen SLNs. Methods: 46 hematoxylin-eosin (HE)-negative sentinel node samples from 42 patients with malignant melanoma were included in this study. Formalin-fixed paraffin-embedded sections were immunostained with S-100 protein and HMB-45. A central portion of the node was submitted for PCR. This method was accomplished with a combination of reverse transcription and amplification of the tyrosinase complementary DNA and double-round PCR (nested reverse transcriptase [RT]-PCR). Results: In 1 of the 42 SLN-negative patients, immunohistochemistry stains allowed the detection of micrometastases. With molecular biology, 14 of the 42 SLN patients were positive (33%); in another 12 (29%), only the nested RT-PCR was positive. Of the 42 patients, 24 were put into 3 groups and followed for a 5-year period with 1, 7, and 16 patients, respectively, in the groups. The first group involved 1 patient who had provided 2 SLN samples that were found to be SLN-positive using both techniques, immunohistochemistry stains and nested RT-PCR (he had hepatic metastasis and died 24 months after diagnosis). The second group, with only nested RT-PCR positive SLN samples, included 7 of 12 patients who were followed and had a median survival of 37 months; 4 died of widespread metastatic disease, the other 3 patients had event-free survival, but 1 consented to undergo a therapeutic lymphadenectomy as a result of a positive test. The last group consisting of 16 of 32 patients, with complete 5-year survival, who were SLN-negative with both techniques, immunohistochemistry stains and nested RT-PCR. Fourteen of the 16 (88%) were event-free survival during the follow-up, and 2 had local relapse. Conclusion: Tyrosinase mRNA amplification may be a negative prognostic factor for the detection of micrometastases in fresh frozen SLNs using molecular biology techniques. © 2004 Adis Data Information BV. All rights reserved.