Objective: Serum S-100B is a reliable tumor marker of malignant melanoma, but efficient use is restricted to patients with metastatic disease. Therefore, the aim of our study was to assess serum S-100B levels at different stages of malignant melanoma and to compare these levels with the expression of the S-100B phenotype in primary tumors and lymph node metastases. Methods: Fifty-nine patients were included in this study; serum S-100B protein was measured using an immunoluminometric assay while the expression pattern in the primary tumor was determined by immunohistochemistry using an anti-S-100B monoclonal antibody. Results: Serum S-100B concentrations were significantly elevated in stage III (p = 0.01) patients, with normal levels in stage I–II. The most frequent S-100B protein expression pattern of the melanoma tissue was found to be diffuse staining observed in around half of the cases (52.5%) followed by heterogeneous (30.5%) and focal patterns (17%), being independent of the stage as well as the lymph node involvement. In stage I–II patients, the various staining patterns did not correlate with the serum concentration of the S-100B protein, while in stage III patients with heterogenous or diffuse S-100B staining patterns in tumor tissue, the serum marker concentration was significantly higher (p < 0.05) than in patients with focal staining. Furthermore, S-100B staining of the melanoma tissue also differed (low/negative, medium and strong staining), and serum marker concentrations corresponded to the pattern of the staining intensity. In stage I–II, only strong staining was associated with elevated serum S-100B concentrations while in stage III medium and strong staining was found to be associated with significantly higher serum marker concentrations compared to patients with tumors with low/negative staining (p < 0.05). Conclusions: In malignant melanoma characterized by focal and/or low S-100B staining in the tumor tissue determined by immunohistochemistry, S-100B monitoring in the serum may not suffice to detect disease progression.