The aim of this study was to determine whether excision biopsy and primary closure of primary cutaneous melanoma modifies lymphatic drainage and accuracy of sentinel node biopsy. Thirty patients with 31 cutaneous melanomas were prospectively enrolled to undergo lymphoscintigraphy (LS) before and after excision biopsy. 99mTc-human serum albumin nanocolloid was first injected intradermally around the primary tumor and subsequently, after excision biopsy, adjacent to the scar. Sentinel nodes were identified by preoperative LS and the γ-probe. Patent Blue V dye was injected intraoperatively before sentinel node biopsy. Intraoperative sentinel node identification was 100%. In 23 of 31 cases, both LSs were concordant in terms of nodal basins visualized. Two patients had a basin downstaged and six patients had a basin upstaged by the second LS. Only 50% of LS hot nodes stained blue (42 of 84). In 24 of 31 cases, the sentinel node was negative for metastases. Seven patients underwent complete lymph node dissection because of sentinel node positivity. Only one patient had metastases also to a non-sentinel node. After a median follow-up of 30 months lymph node metastases have not been observed in the eight discordant cases. This study shows that sentinel node identification and biopsy after lymphatic mapping is accurate after excision biopsy of primary cutaneous melanoma. Excision biopsy may, however, modify lymphatic drainage and a narrow excision margin should be performed if melanoma is suspected.