The objective of this study was to evaluate the role of preoperative 18F-fluorodeoxyglucose-positron emission tomography/computed tomography scanning, preoperative lymphoscintigraphy (LS), and sentinel lymph node biopsy in patients with malignant melanoma. Fifty-two patients (36 men: 16 women; mean age 55.0+/-13.0 years; median age 61 years; range 17-76 years) with malignant melanoma were selected. According to the latest version of the American Joint Committee on Cancer staging system, the disease in the study patients was initially classified as either stage I or II. The other primary tumor characteristics were mean Breslow depth=2.87 mm and median=2 mm; range 1-12.0 mm and Clarks levels III-V. None of the study patients had clinical or radiological evidence of regional lymph node metastatic disease. At least one sentinel node was identified in all patients. Preoperative LS detected a total of 111 sentinel lymph nodes (average 2.13 sentinel lymph node per patient) and demonstrated a single nodal draining basin in 38 (73%) patients and multiple (2-3 draining basins) in the remaining 14 (27%) patients. Fourteen out of the 52 patients (27%) had at least one involved sentinel node. Positron emission tomography was true positive in two patients with a sentinel node greater than 1 cm and false positive in two other patients. In this study, the detection of sentinel lymph node by LS and gamma probe had a sensitivity of 100%. In contrast, 18F-FDG-PET imaging demonstrated very low sensitivity (14.3%; 95% CI, 2.5 to 44%) and positive predictive value (50%; 95% CI, 9 to 90%) for localizing the subclinical nodal metastases. The specificity, net present value, and diagnostic accuracy were 94.7, 75, and 73%, respectively. Preoperative fluorodeoxyglucose-positron emission tomography/computed tomography imaging is not able to substitute LS/sentinel lymph node biopsy in patients at stage I or II.