BRAF inhibitors have demonstrated improvement of overall survival in patients with metastatic melanoma and BRAF(V600) mutations. In order to evaluate BRAF tumor heterogeneity between primary and metastatic site, we have evaluated the performance of immunohistochemistry (IHC) with an anti-BRAF(V600E) antibody in both localization by comparison with high resolution melting analysis followed by Sanger sequencing in a parallel blinded study. A total of 230 samples distributed as primary melanoma (n = 88) and different types of metastatic samples (n = 142) were studied in 99 patients with advanced or metastatic melanoma (stage III or IV). The prevalence of each BRAF mutation was c.1799T>A, BRAF(V600E) (45.2%), c.1799_1800TG>AA, BRAF(V600E2) (3.0%), c.1798_1799GT>AA, BRAF(V600K) (3.0%), c.1801 A>G, BRAF(K601E) (1.3%), c.1789_1790CT>TC, BRAF(L597S) (0.4%), c.1780G>A, BRAF(D594N) (0.9%) respectively. IHC was positive in 109/112 samples harboring BRAF(V600E/E2) mutations and negative in other cases. The cytoplasmic staining was either strongly positive in tumor cells of BRAF(V600E) mutated cases. It appeared strong brown, different from the vesicular grey cytoplasmic pigmentation of melanophages. Concordance between the two techniques was 96.4%. Sensitivity of IHC for detecting the BRAF(V600E/E2) mutations was 97.3%, while specificity was 100%. Both our IHC and molecular study demonstrated homogeneity between primary and metastatic sites for BRAF status in melanoma. This study also provides evidence that IHC may be a cost-effective first-line method for BRAF(V600E) detection. Thereafter, molecular techniques should be used in negative, ambiguous or non-contributive cases.