Background: Sentinel lymph node biopsy (SLNB) is an important tool for accurate staging of patients with melanoma. There is an ongoing debate whether the procedure provides therapeutic benefits or not. Objective: We wanted to analyze 10-year data from an academic teaching hospital. Patients and methods: During 1999-2009, 977 patients with early cutaneous melanoma have been treated. Of these, 419 patients had tumors ≥1 mm thickness. Patients with head and neck tumors were excluded, leaving 364 patients (202 men and 162 women). SLNB was not performed in 163 patients but was performed in 201 patients. For correction of bias, tumors >4 mm thickness were excluded from further statistical analysis. Results: The detection rate of SLN was 94.4%. False negative SLN were observed in 8.9%. Adverse effects occurred in 5.5%. The rate of positive SLNB was 16.4% and lymph node involvement 20%. Patients undergoing SLNB had a lower relapse rate (10.6% vs. 33.3%; P < 0.001). The most important finding is an almost 50% lower total death rate and melanoma-related death rate in the SLNB subgroup (P < 0.001 for both). Conclusions: Sentinel lymph node biopsy is an accurate instrument for melanoma staging to detect occult regional lymph node involvement. Data suggest a positive effect on relapse-free survival. The observation of improved long-term survival needs validation in prospective multicenter trials. The limitations of this study were that it was a single center retrospective analysis.